fv^j i grr 3 states 1 · 2014-06-23 · fv igrr unitedstatesdepartmentofcommerce...
TRANSCRIPT
fV IGrr United states Department of Commerce1 ^J In 3 1 National Institute of Standards and Technology
NATL INST. OF STAND S TECH R.I.C.
Illl llll ill,
A111D3 M5DD7b
NIST
PUBLICATIONS
NIST Technical Note 1265
Guidelines for Realizing the
International Temperature Scale of
1990 (ITS-90)
B. W. Mangum and G. T. Furukawa
QC
100
.U5753
#1265
1990
C.2
NATIONAL INSTITUTE OF STANDARDS &TECHNOLOGY
Research Informatkm CenterGakhersburg, MD 20899
146753
HIST Technical Note 1265
Guidelines for Realizing the
International Temperature Scale of
1990 (ITS-90)
B. W. MangumG. T. Furukawa
Center for Chemical TechnologyNational Measurement Laboratory
National Institute of Standards and TechnologyGaithersburg, MD 20899
August 1990
SfATE5 O**
U.S. Department of CommerceRobert A. Mosbacher, Secretary
National Institute of Standards and Technology
John W. Lyons, Director
National Institute of Standards U.S. Government Printing Office For sale by the Superintendent
and Technology Washington: 1 990 of DocumentsTechnical Note 1265 U.S. Government Printing Office
Natl. Inst. Stand. Technol. Washington, DC 20402Tech. Note 1265
190 pages (Aug. 1990)
CODEN: NTNOEF
ACKNOWLEDGMENTS
The authors gratefully acknowledge the assistance of Mr. Gregory F. Strouse inproviding some of the figures and tables and the reports of calibration forSPRTs, and the assistance of Mr. George W. Burns, Mr. Earl R. Pfeiffer and MissJacquelyn A. Wise in providing copies of reports of calibration forthermocouples, RIRTs , and liquid- in- glass thermometers, respectively.
111
Table of Contents
page
List of Tables ix
List of Figures xAbstract xiiiScope xiii1
.
Introduction 1
2. Definition of the ITS- 90 3
2.1 Between 0.65 K and 5.0 K ,
3He and 4He vapor-pressurethermometry 6
2.2 Between 3.0 K and 24.5561 K (the triple point of Ne)
,
3He and AHe constant volume gas thermometry 10
2.3 Between 13.8033 K (the triple point of equilibriumhydrogen) and 1234.93 K (the freezing point of silver),platinum resistance thermometry 11
2.3.1 General relation between resistance ratiosand Tgo 11
2.3.2 SPRT specifications 12
2.3.3 Range 13 . 8033 K to 273 . 16 K 14
2.3.3.1 Subrange 24.5561 K to 273.16 K 16
2.3.3.2 Subrange 54.3584 K to 273.16 K 16
2.3.3.3 Subrange 83.8058 K to 273.16 K 16
2.3.4 Range 273.15 K (0 °C) to 1234.93 K (961.78 °C) 17
2.3.4.1 Subrange 273.15 K (0 °C) to 933.473 K(660.323 °C, freezing point ofaluminum) 18
2.3.4.2 Subrange 273.15 K (0 °C) to 692.677 K(419.527 °C, freezing point of zinc) 18
2.3.4.3 Subrange 273.15 K (0 °C) to 505.078 K(231.928 °C, freezing point of tin) 18
2.3.4.4 Subrange 273.15 K (0 °C) to 429.7485 K(156.5985 °C, freezing point ofindium) 18
2.3.4.5 Subrange 273.15 K (0 °C) to 302.9146 K(29.7646 °C, melting point ofgallium) 19
2.3.5 Subrange 234.3156 K (-38.8344 °C, triple point ofmercury) to 302.9146 K (29.7646 °C, the meltingpoint of gallium) 19
2.4 Above 1234.93 K (961.78 °C, freezing point of silver);radiation thermometry based on Planck's Law ofRadiation 19
3 . Realization of the ITS - 90 20
3.1 Vapor pressure thermometry and gas thermometry 203.1.1 Realization of the ITS -90 below 273.16 K 20
3.1.2 Vapor pressure thermometry and the CVGT range 203.1.3 Realization of the vapor-pressure and CVGT scales
at temperatures below the neon triple point 22
3.1.3.1 3He vapor pressure measurements 233.1.3.2 AHe vapor pressure measurements 24
iv
Table of Contents (continued)
page
3.1.3.3 e-H2 vapor pressure measurements 24
3.1.3.4 Constant volume gas thermometry 24
3.1.3.5 Pressure measurements 26
3.2 Realization of the fixed points of the ITS-90 28
3.2.1 Effect of impurities 28
3.2.2 Triple points of e-H2 , Ne , 2 , and Ar 29
3.2.2.1 General consideration of apparatusdesign 29
3.2.2.2 Realizations of the triple points andtheir application to calibration 33
3.2.3 Triple point of water, 273.16 K (0.01 °C) 36
3.2.3.1 Realization and application of thetriple point of water 37
3.2.4 Freezing, melting, or triple points of metals: Hg,
Ga, In, Sn, Zn, Al , Ag, Au, or Cu 39
3.2.4.1 Container material 42
3.2.4.2 Metal fixed point devices forcalibrating SPRTs 42
3.2.5 Control charts of check thermometers 59
3 .
3
Radiation thermometry 60
4. Calibration of thermometers on the ITS-90 at various levels ofuncertainty and some approximations of the scale 61
4.1 Rhodium- iron resistance thermometers 63
4.2 Germanium resistance thermometers 64
4.3 Standard platinum resistance thermometers 64
4.3.1 Capsule SPRTs [13.8033 K to 429.7485 Kor 505.078 K] 64
4.3.2 Long-stem type SPRTs (83.8058 K to 1234.93 K) 64
4.3.3 Conversion of the IPTS-68 constants and W(T6Q )
tables to approximate ITS-90 constants andW(TQQ ) tables 65
4.3.4 Uncertainties of calibrations and theirpropagation 69
4.3.5 Estimates of possible errors introduced byextrapolations beyond the range of calibration 70
4.4 Thermocouples (77 K to 2400 K) 78
4.5 Liquid- in- glass thermometers 78
4.6 Industrial platinum resistance thermometers 78
4.7 Thermistor thermometers, digital thermometers, and othertypes of thermometers 79
4.8 The logo of the National Conference of StandardsLaboratories for the ITS-90 80
5 References 82
Table of Contents (continued)
page
Appendices 90
6.1 Copies of, or excerpts from, some papers relatedto the work of the international bodies concernedwith the ITS - 90 90
6.1.1 Reproduction of the article by H. Preston-Thomas,entitled, "The International Temperature Scale of1990," published in Metrologia 27, 3-10 (1990),with corrections 90
6.1.2 Reproduction of the article by T. J. Quinn,entitled "News from the BIPM"
,published in
Metrologia 26, 69-74 (1989) 99
6.1.3 Recommendations adopted by the CCT at its
17th Session 106
6.1.3.1 Recommendation Tl (1989), TheInternational Temperature Scale of 1990 106
6.1.3.2 Recommendation T2 (1989), Reference tablesfor thermocouples and industrial platinumresistance thermometers 107
6.1.3.3 Recommendation T3 (1989), The uncertaintyinherent in the realization of theInternational Temperature Scale of 1990 107
6.1.4 Reproduction of article by B. W. Mangum,entitled "Special Report on the InternationalTemperature Scale of 1990; Report on the 17thSession of the Consultative Committee onThermometry," published in J. Res. Natl. Inst.
Stand. Technol. 95, 69-77 (1990) 108
6.2 Excerpts from proceedings of some meetings of theCIPM and of the CGPM that are related to the designationof the size of the kelvin and to the ITS -90 118
6.2.1 Recommendation of the CIPM that the numerical valueof 273.16 °K, exactly, be assigned to the triple pointof water on the thermodynamic scale, excerpted fromProces-Verbaux des Seances, Deuxieme Series, Tome XXIV,Session de 1954 du Comite International des Poids etMesures, published by Gauthier Villars , Editeur-Imprimeur-Libraire , Paris , 1955 118
6.2.2 Definition of the thermodynamic scale of temperature.This entails the assignment of the value 273.16 °K,
exactly, to the triple point of water as afundamental fixed point, excerpted from Comptes Rendusdes Seances, Dixieme Conference Generale des Poids etMesures, Reunie a Paris en 1954, published by GauthierVillars, Editeur-Imprimeur-Libraire , Editeur du BureauInternational des Poids et Mesures, Paris, 1955 118
VI
Table of Contents (continued)
page
6.2.3 Recommendations of the CCT, concerning thedevelopment of the ITS-90 and simplified methodsof its realization, that were presented to the
CIPM on 9 July 1987, excerpted from Proces-Verbauxdes Seances du Comite International des Poids et
Mesures , Proces-Verbaux de la 76 e session - 1987,
Tome 55, Edite par le BIPM, Pavilion de Breteuil,F- 92312 Sevres Cedex, France 119
6.2.4 The resolution, instructing the CIPM to develop theITS-90, adopted by the 18th CGPM in 1987, and excerptedfrom Comptes Rendus , 18 e Conference Generale des Poidset Mesures (1987), Edite par le BIPM, Pavilion de
Breteuil, F-92312 Sevres Cedex, France 1216.3 Examples of some typical Report of Calibration documents
from the NIST Calibration Laboratories for thermometerscalibrated over various ranges 122
6.3.1 Example of a Report of Calibration for an SPRTcalibrated over the range from 13.8033 K to
302 . 9146 K 122
6.3.2 Example of a Report of Calibration for an SPRTcalibrated over the range from 83.8058 K to
933.473 K 1266.3.3 Example of a Report of Calibration for an SPRT
calibrated over the range from 273.15 K to
1234.93 K 129
6.3.4 Example of a Report giving the T68 to T90
conversion for a long- stem type SPRT calibratedon the IPTS-68(75) from the oxygen condensationpoint to the freezing point of zinc 132
6.3.5 Example of a Report giving the T68 to r90 conversionfor a capsule -type SPRT calibrated on the IPTS-68(75)from the triple point of equilibrium hydrogen to
the freezing point of tin 135
6.3.6 Example of a Report of Calibration for an RIRTcalibrated over the range from 0.50 K to 30 Kon the EPT-76 138
6.3.7 Example of a Report of Calibration for an RIRT(the RIRT of sec. 6.3.6) calibrated over therange from 0.65 K to 25 K on the ITS-90 147
6.3.8 Example of a Report of Calibration for a thermocouplecalibrated over the range from °C to 1450 °C 154
6.3.9 Example of a Report of Calibration for a
liquid- in- glass thermometer calibrated over therange from °C to 100 °C on the IPTS-68(75) 162
VII
Table of Contents (continued)
page
6.3.10 Example of a Report of Calibration for theliquid- in- glass thermometer of section 6.3.9calibrated over the range from °C to 100 "C onthe IPTS-68(75) but converted to the ITS-90 164
6.4 The National Conference of Standard Laboratories (NCSL)Ad Hoc Committee (91.3) on the Change in the InternationalTemperature Scale that was formed to publicize the ITS-90and its implementation and to minimize the confusion andinconvenience of the change from the IPTS-68(75) 166
7 . NIST Calibration Services , Thermometers 167
7 .
1
Laboratory Thermometers 167
7 .
2
Thermocouples , Thermocouple Materials , andPyrometer Indicators 169
7 .
3
Resistance Thermometers 173
vm
List of Tables
page
Table 1. Differences between r90 and T68 (and t90 and t68 ) , andbetween T90 and T76 5
Table 2 . Defining fixed points of the ITS -90 8
Table 3. Comparison of temperatures of fixed points assigned onvarious scales 9
Table 4. Values of the coefficients Atand of the constants B and C
for the 3He and AHe vapor pressure equations and the temperaturerange for which each equation is valid 10
Table 5. Values of the coefficients Ai? B A , C tand D L and of the
constants A , B , C , and D in the reference functions,equations (14) and (22) , and in the functions approximatingthem, given by equations (15) and (23) 15
Table 6 . The effect of pressure on the temperatures of the definingfixed points. The reference pressure for the equilibriumstates of freezing and melting points is one standardatmosphere (101,325 Pa). Triple points have the vapor pressureof the material when the solid, liquid and vapor phasesare present in equilibrium 25
Table 7. Latent heats of fusion and first cryoscopic constants ofdefining fixed-point materials 29
Table 8. Example of conversion of calibrations of SPRTs on the IPTS-68to approximate calibrations on the ITS -90 66
Table 9. Values of P/(T68 ) and W(TQQ ) for various fixed-point temperaturesT68 and T90 , respectively 67
Table 10. Example of a conversion of calibration values of a type Kthermocouple on the IPTS-68 to an approximate calibrationon the ITS -90 79
IX
List of Figures
Page
Figure 1. The temperature difference (t90 - t68 )/°C in therange between the triple point of equilibriumhydrogen (-259.3467 °C) and the freezing point ofgold (1337 . 33 °C) 4
Figure 2. A schematic representation of the ITS -90 showingthe temperatures of the defining fixed points (or
phase equilibrium states) on the scale and temperatureranges defined by interpolation instruments and equations ... 7
Figure 3. A schematic representation of the ITS -90 in the rangespecified for the platinum resistance thermometer,showing the various defined subranges and the
temperatures of the defining fixed points on the scalerequired for calibration 13
Figure 4. A schematic drawing of the NIST argon triple -pointapparatus for calibrating seven long- stem SPRTs andsix capsule SPRTs 30
Figure 5 . A sealed cell suitable for containing cryogenic gasesat high pressures 31
Figure 6. Two types of triple point of water cells with wellsfor platinum resistance thermometers 35
Figure 7 . An SPRT in a Type A triple point of water cell immersedin an ice bath 38
Figure 8. Idealized liquid/solid equilibrium conditions inside fixedpoint cells used in freezing and melting experiments 40
Figure 9 . An SPRT in a metal freezing-point cell 41
Figure 10. Two mercury triple -point cells, one constructed ofborosilicate glass and one of Type 304 stainless steel 43
Figure 11. A borosilicate glass, mercury triple-point cell in a
stainless steel container 45
Figure 12. Photograph of an all-plastic indium cell and its
stainless steel container 46
Figure 13. An all-plastic gallium melting/triple-point cell 48
x
List of Figures (continued)
Page
Figure 14. An all-plastic indium freezing-point cell to be usedin a stainless steel container, such as that shownin figure 12 50
Figure 15. A graphite freezing point cell enclosed inside a
fused silica tube with tube connection to highvacuum, purified argon gas source, and pressure gauge 53
Figure 16. A method for filling a graphite freezing-point cell bymelting the metal sample in the graphite crucible 55
Figure 17. An apparatus for installing a graphite thermometerwell and lid in a graphite crucible containing a
molten metal sample 56
Figure 18. Differences between r90 and (r90 ) as calculated from IPTS _ 68 calibrationfor the long-stem SPRT, S/NRS8YA-5, of table 8 68
Figure 19. Propagation of errors from errors of calibration ofSPRTs between 13 . 8033 K and 273 . 16 K 71
Figure 20. Propagation of errors from errors of calibration ofSPRTs between 273.15 K and 1234.93 K. Also includedin this figure are error curves for errors madeby the user at the triple point of water 72
Figure 21. Curve for a NIST SPRT that shows the error introducedby extrapolating the deviation function, determinedfrom calibration over the range from the triple pointof argon to the triple point of water, downward fromthe triple point of argon to 54 K 73
Figure 22. Curve for a NIST SPRT that shows the error introducedby extrapolating its deviation function, determinedfrom calibration over the range from the triple pointof mercury to the melting point of gallium, downwardfrom the triple point of mercury to 84 K 74
Figure 23. Curve for the NIST SPRT of figure 22 that shows the errorintroduced by extrapolating its deviation function,determined from calibration over the range from thetriple point of mercury to the melting point of gallium,downward from the triple point of mercury to only 200 K 75
XI
List of Figures (continued)
Page
Figure 24. Curves for several NIST SPRTs that show the errorsintroduced by extrapolating their deviation functions,determined from calibration over the range from thetriple point of water to the freezing point of zinc,downward from the triple point of water to -50 °C 76
Figure 25. Curves for several NIST SPRTs that show the errorsintroduced by extrapolating their deviation functions
,
determined from calibration over the range from thetriple point of water to the freezing point of zinc,upward from the freezing point of zinc to 934 K (660 °C)
.
Also shown are subrange inconsistencies for the subrangetriple point of water to zinc, relative to the subrangetriple point of water to aluminum 77
Figure 26 . The NCSL ITS -90 logo 81
XII
ABSTRACT
This Technical Note describes the International Temperature Scale of 1990
(ITS-90) that became the official international temperature scale on 1 January
1990, superseding the previous scales, and provides information on how the scale
may be realized at different levels of accuracy. The ITS-90 extends upward from0.65 K, is in close agreement with the Kelvin Thermodynamic Temperature Scale,
has much improved continuity, precision and reproducibility throughout its rangesover that of previous scales, and has subranges and alternative definitions in
certain ranges that greatly facilitate its use. In addition to a descriptionof the ITS-90 and how it can be realized, there are included in this documentreproductions of some articles and excerpts from documents concerned with the
ITS-90. The composition of the Comite Consultatif de Thermometrie (CCT) of the
Comite International des Poids et Mesures (CIPM) at the time of the adoption ofthe ITS-90 is given. The differences between the temperatures on the ITS-90 andthose on the International Practical Temperature Scale of 1968, Amended Editionof 1975, [IPTS-68(75)] and those on the 1976 Provisional 0.5 K to 30 KTemperature Scale (EPT-76) are tabulated. Measurement procedures for realizingthe ITS-90 throughout the various ranges of the scale are given. Also, for the
most important temperature region, the region of the platinum resistancethermometer (PRT) , computational examples are given for determining the
coefficients of the relevant deviation equations for PRTs calibrated at varioussets of fixed points. The effects of the introduction of the ITS-90 onelectrical reference standards are addressed also.
SCOPE
The ITS-90 and, consequently, the topics discussed in this document affect allaspects of thermometry at temperatures from 0.65 K upward. This change of scalewill affect principally those making high precision temperature measurements.Thus, all standards and calibration laboratories will be affected and they shouldimplement the ITS-90 as soon as possible.
If the required minimum uncertainty of measurements over a given range oftemperature is at least three times as large as the differences between the newand old scales in that range, as given in table 1 and shown in figure 1 of thisdocument, then the effect of the change in scales will be negligible and the userwill, for all practical purposes, be unaffected.
DISCLAIMER
Certain commercial equipment, instruments, or materials are identified in thispaper in order to adequately specify the experimental procedure. Suchidentification does not imply recommendation or endorsement by the NationalInstitute of Standards and Technology, nor does it imply that the materials orequipment identified are necessarily the best available for the purpose.
Xlll
Guidelines for Realizing the
International Temperature Scale of 1990 (ITS -90)
B . W . MangumNational Institute of Standards and Technology
Gaithersburg, MD 20899
and
G. T. FurukawaGuest Scientist
National Institute of Standards and TechnologyGaithersburg, MD 20899
1 . INTRODUCTION
The Comite Consultatif de Thermometrie (CCT) is one of eight specializedtechnical subcommittees of the Comite International des Poids et Mesures (CIPM)
.
The CIPM is a committee of the Conference Generale des Poids et Mesures (CGPM)
.
These eight subcommittees are:
1. The Comite Consultatif d' Electricite (CCE) , established in 1927,
2. The Comite Consultatif de Photometrie et Radiometric (CCPR) , assignedthis name in 1971; the previous name was the Comite Consultatif de
Photometrie, established in 1933,
3. The Comite Consultatif de Thermometrie (CCT), established in 1937,4. The Comite Consultatif pour la Definition du Metre (CCDM) , established
in 1952,5. The Comite Consultatif pour la Definition de la Seconde (CCDS)
,
established in 1956,6
.
The Comite Consultatif pour les Etalons de Mesure des RayonnementsIonisants (CCEMRI) , established in 1958,
7. The Comite Consultatif des Unites (CCU) , established in 1964, and8. The Comite Consultatif pour la Masse et les grandeurs apparentees
(CCM), established in 1980.
The CCT is composed presently of members from the following laboratories:1. Amt fur Standardisierung, Messwesen und Warenpriifung [ASMW] , Berlin,
DDR,
2. Bureau National de Metrologie, Paris, France : Institut National de
Metrologie [INM], du Conservatoire National des Arts et Metiers,3. Ceskoslovensky Metrologicky Ustav [CSMU] , Bratislava, Czechoslovakia,4. National Research Council [NRC] , Ottawa, Canada,5. CSIRO, Division of Applied Physics [CSIRO], Lindfield, Australia,6. D.I. Mendeleyev Institute for Metrology [VNIIM] , Leningrad, USSR;
Physico- Technical and Radio -Technical Measurements Institute [PRMI]
,
Moscow, USSR,7. National Institute of Metrology [NIM] , Beijing, PRC
,
8. Istituto di Metrologia G. Colonnetti [IMGC] , Turin, Italy,
9. Kamerlingh Onnes Laboratorium [KOL] , Leiden, The Netherlands,10. Korea Standards Research Institute [KSRI], Seoul, Korea,
11. National Institute of Standards and Technology [NIST] , Gaithersburg,MD, USA,
12. National Physical Laboratory [NPL] , Teddington, UK,
13. National Research Laboratory of Metrology [NRLM] , Ibaraki, Japan,14. Physikalisch-Technische Bundesanstalt [PTB] , Braunschweig, FRG,
15. Van Swinden Laboratorium [VSL] , Delft, The Netherlands,16. Iowa State University, Ames, Iowa, USA, and17. Bureau International des Poids et Mesures [BIPM] , Sevres, France.
Shortly after the adoption of the International Practical Temperature Scale of1968 (IPTS-68) [100] , it was realized that the scale had many deficiencies andlimitations. These included its lower limit of 13.81 K, its inaccuracy relativeto thermodynamic temperatures, and its non-uniqueness and irreproducibility
,
especially in the temperature region from T68 = 903.89 K (630.74 °C) to
T68 = 1337.58 K (1064.43 °C) , the region in which the Pt-10%Rh/Pt thermocouplewas the standard interpolating instrument. Consequently, the CCT undertook the
development of a new scale, and completed it in accordance with Resolution 7 ofthe 18th Conference Generale des Poid et Mesures [29] , which met in October 1987(see appendices)
.
The CCT met 12-14 September 1989 at the Bureau International des Poids et Mesures(BIPM) in its 17th Session [14] and completed the final details of the newtemperature scale, the International Temperature Scale of 1990 (ITS-90) [66,83].The CCT then recommended to the CIPM, which met [84] on 26-28 September 1989 atthe BIPM, that the ITS-90 be adopted and made the official scale (see
appendices) . Upon considering this recommendation, the CIPM adopted the newtemperature scale (see appendices) , and the ITS-90 became the officialinternational temperature scale on 1 January 1990, the same date on which changesaffecting certain electrical reference standards were implemented [12]. TheITS-90 supersedes the IPTS-68, the International Practical Temperature Scale of1968, Amended Edition of 1975 [IPTS-68(75)
] [101] , and the 1976 Provisional 0.5 Kto 30 K Temperature Scale (EPT-76) [99].
The ITS-90 was implemented at the NIST on 1 January 1990. The purpose of thisdocument is to describe the new scale, to give some guidelines for its
realization and use, to facilitate its implementation, to give the differencesbetween temperatures on it and those on the IPTS-68(75) and on the EPT-76, andto describe how the NIST realizes the scale.
The ITS-90 extends upward from 0.65 K and temperatures on this scale are in muchbetter agreement with thermodynamic values than are those on the IPTS-68 (75) andthe EPT-76. The new scale has subranges and alternative definitions in certainranges that greatly facilitate its use. Furthermore, its continuity, non-uniqueness and reproducibility throughout its ranges are much improved over thecorresponding characteristics of the previous scales. The biggest improvementin reproducibility results from the replacement of thermocouple thermometry withplatinum resistance thermometry in the range 630 °C to the freezing-pointtemperature of silver, and with radiation thermometry in the range from thefreezing-point temperature of silver to that of gold.
The change in the temperature scale affects not only technical interests involveddirectly in thermometry but also those involved with other reference standards,
such as electrical standards, that are sensitive to temperature. As examples,standard resistors and standard cells are sensitive to temperature and generallyare maintained in constant- temperature environments, at least in nationalstandards laboratories. At the present time, the temperatures of thoseenvironments are normally determined with thermometers that were calibrated onthe IPTS-68(75). A given thermodynamic temperature expressed on the ITS-90,
however, has a value that is different from that expressed on the IPTS-68(75),except at absolute zero (0 K) , at the triple-point temperature of water(273.16 K) , and at a few other points at which the temperatures on the two scalesare fortuitously the same. This difference is shown in figure 1 [83]. A tableof differences between temperatures on the ITS-90, i.e. , r90 or t90 , and those onthe IPTS-68(75), i.e., r68 or t68 , and those on the EPT-76, T76 , is given in the
text of the ITS-90 and is presented here in table 1. Although temperature valuesexpressed on the two scales are different, the change is only in the expressionof the value of temperature and not in the temperature itself . That is to say,
the Kelvin thermodynamic temperature (the hotness) of a material at any givenpoint is independent of the use of any of the 'practical' temperature scales.The better the 'practical' scale is, the closer the values of temperature on thatscale are to the thermodynamic temperature values. Needless to say, the Kelvinthermodynamic temperature values are experimentally determined, and they may havesignificant error. Since temperature values expressed on the thermodynamic and'practical' scales are different, if the temperature of the environment of areference standard is adjusted so that its value when expressed on the ITS-90has the same value as had been used on the IPTS-68(75), there will have been achange of the thermodynamic temperature and the value of the reference standardwill usually change. Of course, one may not want to change the thermodynamictemperature of the reference standard. In that case, the thermodynamictemperature, as expressed on the IPTS-68(75), can simply be expressed on the ITS-
90 (a numerical value different from that on the IPTS-68(75)) and the referencestandards will be unaffected. For more details on the effects of the change ofthe temperature scale on electrical standards, see National Institute ofStandards and Technology (NIST) Technical Note 1263 [12].
In addition to the effect on reference standards for measurements, alltemperature -sensitive properties, e.g., tables of thermodynamic properties, thatare presently expressed on the IPTS-68(75) may require changes in values.
2. DEFINITION OF THE ITS-90
The ITS-90 was designed by the CCT in such a manner that temperature valuesobtained on it do not deviate from the Kelvin thermodynamic temperature valuesby more than the uncertainties of the latter values at the time the ITS-90 wasadopted. Thermodynamic temperature is indicated by the symbol T and has the unitknown as the kelvin, symbol K. The size of the kelvin is defined to be 1/273.16of the thermodynamic temperature of the triple point of water. This definitionof the Kelvin Thermodynamic Temperature Scale (KTTS) that is based on the valueof a single finite temperature is not new; the CCT proposed it in 1954, the CIPMrecommended it, and the Tenth CGPM adopted it that same year [30]
.
Because temperatures on previous temperature scales were expressed relative to
the ice point (273.15 K) , temperature, symbol t, on the Celsius Temperature Scaleis defined by:
•Hu
cTeg
oCM
o o-•Hi-l
•i-l
cr
1
1 1 I 1 1 1 / 1
o a)
o 4-i
o o
4->
c•HO
o0)
i-H
P.•i-l
o00
the
tr
85].
o
m COQ) 00
4J
-
o
o
oao
the
range
b(
(1064.18
°C
— o o O0)
r> "O
1 1 '/ o/ +J U 00
o
o o
^ oCOID 1}
w c•H
1 O— — o — o o &
en
CM
J
CM
e
(t,
zing
/O Q>
/o. CD !-J
}-l 4-1
(1)
°4-1 <D
4-1 43.,-1 4->
— '* ^~^^^ — O ' — o73
Tl"D0) CJ-l cfl
3
O °t! oZ <D
to, r^t- CM E v£>
1 1 I
oCM
1
/ o o —o" o
1 1
1 1 1
oCM
1
he
te
59.34
w 1 i w 1 1 1
1.
T
2n
(-2\\
CMoo
O CMOo
oo
o NcT
1
1 1 0) b0^ O3 H
4
3o/(89J-.06 l) 90U9J9^!P 9.in \Bii)diu91
bOT)
Table 1. Differences between T90 and r68 (and tgo and t68 ) , and between r90 and T
(T90 - r76)/mK
-0.1 -0.2 -0.3 -0.4 -0.5
10 -0.6 -0.7 -0.8 -1.0 -1.1 -1.3 -1.4 -1.6 -1.8 -2.0
20 -2.2 -2.5 -2.7 -3.0 -3.2 -3.5 -3.8 -4.1
<r90 - r68)/K
r90/K 1 2 3 4 5 6 7 8 9
10 -0.006 -0.003 -0.004 -0.006 -0.008 -0.009
20 -0.009 -0.008 -0.007 -0.007 -0.006 -0.005 -0.004 -0.004 -0.005 -0.006
30 -0.006 -0.007 -0.008 -0.008 -0.008 -0.007 -0.007 -0.007 -0.006 -0.006
40 -0.006 -0.006 -0.006 -0.006 -0.006 -0.007 -0.007 -0.007 -0.006 -0.006
50 -0.006 -0.005 -0.005 -0.004 -0.003 -0.002 -0.001 0.000 0.001 0.00260 0.003 0.003 0.004 0.004 0.005 0.005 0.006 0.006 0.007 0.00770 0.007 0.007 0.007 0.007 0.007 0.008 0.008 0.008 0.008 0.00880 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.00890 0.008 0.008 0.008 0.008 0.008 0.008 0.008 0.009 0.009 0.009
r90/K 10 20 30 40 50 60 70 80 90
100 0.009 0.011 0.013 0.014 0.014 0.014 0.014 0.013 0.012 0.012200 0.011 0.010 0.009 0.008 0.007 0.005 0.003 0.001
( C90 " ^68^/ C
tgo/°C -10 -20 -30 -40 -50 -60 -70 -80 -90
-100 0.013 0.013 0.014 0.014 0.014 0.013 0.012 0.010 0.008 0.0080.000 0.002 0.004 0.006 0.008 0.009 0.010 0.011 0.012 0.012
t 90/°c 10 20 30 40 50 60 70 80 90
0.000 -0.002
/
-0.005 -0.007 -0.010 -0.013 -0.016 -0.018 -0.021 -0.024
100 -0.026 -0.028 -0.030 -0.032 -0.034 -0.036 -0.037 -0.038 -0.039 -0.039
200 -0.040 -0.040 -0.040 -0.040 -0.040 -0.040 -0.040 -0.039 -0.039 -0.039
300 -0.039 -0.039 -0.039 -0.040 -0.040 -0.041 -0.042 -0.043 -0.045 -0.046
400 -0.048 -0.051 -0.053 -0.056 -0.059 -0.062 -0.065 -0.068 -0.072 -0.075
500 -0.079 -0.083 -0.087 -0.090 -0.094 -0.098 -0.101 -0.105 -0.108 -0.112
600 -0.115 -0.118 -0.122 -0.125 -0.08 -0.03 0.02 0.06 0.11 0.16700 0.20 0.24 0.28 0.31 0.33 0.35 0.36 0.36 0.36 0.35800 0.34 0.32 0.29 0.25 0.22 0.18 0.14 0.10 0.06 0.03900 -0.01 -0.03 -0.06 -0.08 -0.10 -0.12 -0.14 -0.16 -0.17 -0.181000 -0.19 -0.20 -0.21 -0.22 -0.23 -0.24 -0.25 -0.25 -0.26 -0.26
t90/°C 100 200 300 400 500 600 700 800 900
1000 -0.26 -0.30 -0.35 -0.39 -0.44 -0.49 -0.54 -0.60 -0.66
2000 -0.72 -0.79 -0.85 -0.93 -1.00 -1.07 -1.15 -1.24 -1.32 -1.413000 -1.50 -1.59 -1.69 -1.78 -1.89 -1.99 -2.10 -2.21 -2.32 -2.43
t/°C = T/K - 273.15. (1)
The unit of temperature t is the degree Celsius, symbol °C, and it is bydefinition the same size as the kelvin. A difference of temperature may beexpressed either in kelvins or in degrees Celsius.
Temperatures on the ITS -90 are expressed, in terms of the International KelvinTemperatures, with the symbol T90 , or, in terms of the International CelsiusTemperatures, with the symbol t90 . The unit of the temperature r90 is the kelvin,symbol K, and the unit of the temperature t90 is the degree Celsius, symbol °C.
The relation between T90 and t90 is:
t9o/°C = r90/K - 273.15. (2)
The ITS -90 extends upward from 0.65 K. There are alternative definitions of T90
in certain temperature ranges and they have equal status. In measurements ofthe highest precision made at the same temperature, the alternative definitionswill yield detectable temperature differences. Also, at any given temperaturebetween defining fixed points, different interpolating thermometers that meetthe specifications of the ITS -90 will indicate different temperature values.These latter differences are referred to as the non-uniqueness in the definitionof the ITS -90. The magnitude of the differences that result from these twosources is sufficiently small to be negligible for all practical purposes.
Temperatures on the ITS -90 are defined in terms of equilibrium phase states ofpure substances (defining fixed points) , interpolating instruments , and equationsthat relate the measured property of the instruments to T90 . The equilibriumphase states of the pure substances and the assigned temperatures used as
defining fixed points of the ITS-90 are listed in table 2. Figure 2 showsschematically the defining phase states and temperature ranges defined by the
various interpolating instruments and equations. For comparison purposes, wegive in table 3 the defining fixed points, and their assigned temperatures, ofthe ITS-90 and of all the previous internationally agreed-upon scales.
2.1 BETWEEN 0.65 K AND 5.0 K: 3He and *He VAPOR PRESSURE THERMOMETRY
The ITS-90 is defined between 0.65 K and 3.2 K by the vapor-pressure- temperaturerelation of 3He , and between 1.25 K and 2.1768 K (the A point) and between2.1768 K and 5.0 K by the vapor-pressure- temperature relations of 4He . T90 is
defined by the vapor-pressure equations of the form:
9
T90/K = A + I Ai{[in(p/Pa) - B]/C}\ (3)
i=l
The values of the coefficients AA and of the constants A , B and C of the vapor-pressure equations for 3He and AHe are given in table 4.
— CO
u woa a C2§ re IK
>EE o11 3 C= -o
0) = at m1 .<! X C.*. u ». F
2 «° 2o 3
CO £
re o§ »o 2
3 "o 2o a. m a. 09 ft.
• • O
ft a
„L„„,\,\,
41. cCL.
Q iycni!±— <>-
iN^8 I
x a
DCftOT
reftOS"
«ft._OH
==3==:,
- 5-fiL5- > --0-9-o o
6
£26£za>0
w T)p fi
C ofl
•HO to
P- P*tf 03
<u gK 3•H pIP p
03
W) CC •H•HC C•H O<U •H03 PT3
1-1
a) O,C P-P Ptp 4->
CO
a> J*>
P. ,034-> T3Cfl 03
)-l c03 •l-l
P-.IP
g 03
<u T3P
co
03 <D
,c bOcPcd
bO S-l
•H 03
S"a>
(A
t
O
CO
P
•ucfl
P* O 03
2'
3CT. P.
a2
03
Pa.
EM
T)
i= c
4-)03
OrHcfl
OG CO
•HP
03
Pp ca Oo>
co -^a) 10
u 03
(X P<u cd
5-1 p(0
O•H gp 303 •H6 5-4
OJ ,Q4S •HO r-\
to •P3<03
Q) to
CM CO CCO O
a) x, •-!
54 a w3 t«
bO 5-1 3•H crCn w a)
Table 2. Defining fixed points of the ITS -90
Material 3 Equilib rium Stateb TempT90 (K)
sraturet90 (°c)
Wr (T90 )e
He VP 3 to 5 - 270.15 to
- 268.15
e-H2 TP 13.8033 - 259.3467 0.00119007
e-H2 (or He) VP (or CVGT) » 17 « - 256.15
e-H2 (or He) VP (or CVGT) = 20.3 « - 252.85
Ne c TP 24.5561 - 248.5939 0.00844974
o2 TP 54.3584 - 218.7916 0.09171804
Ard TP 83.8058 - 189.3442 0.21585975
Hgc TP 234.3156 - 38.8344 0.84414211
H2 TP 273.16 0.01 1.00000000
Gac MP 302.9146 29.7646 1.11813889
Inc FP 429.7485 156.5985 1.60980185
Snd FP 505.078 231.928 1.89279768
Zn FP 692.677 419.527 2.56891730
Al c FP 933.473 660.323 3.37600860
Ag FP 1234.93 961.78 4.28642053
Au FP 1337.33 1064.18
Cuc FP 1357.77 1084.62
a e-H2 indicates equilibrium hydrogen, that is, hydrogen with theequilibrium distribution of its ortho and para states at the correspondingtemperatures. Normal hydrogen at room temperature contains 25% para and75% ortho hydrogen.
b VP indicates vapor pressure point or equation; CVGT indicates constantvolume gas thermometer point; TP indicates triple point (equilibriumtemperature at which the solid, liquid and vapor phases coexist) ; FPindicates freezing point and MP indicates melting point (the equilibriumtemperatures at which the solid and liquid phases coexist under a pressureof 101,325 Pa, one standard atmosphere). The isotopic composition is thatnaturally occurring.
c Previously, these were secondary fixed points.
d Previously, these were alternative fixed points.
e From reference functions, eqs (14) and (22).
Table 3. Comparison of temperatures of fixed points assigned on variousscales. Temperatures are expressed in kelvins on the KTTS or equivalent scales
Point NHS a ITS-27b ITS--48b IPTS--48b IPTS-68 IPTS-68(75) EPT -76 ITS-90
Au FPC _ 1336 15 1336 15 1336 15 1337 58 1337.58 _ 1337.33
Ag FP - 1233 65 1233 95 1233 95 1235 08 1235.08 - 1234.93
Al FP - - - - - - - 933.473
S BPd - 717 75 717 75 717 75 - - - -
Zn FP - - - (692 655) 692 73 692.73 - 692.677
Sn FP - - - - (505 1181)(505.1181) - 505.078
In FP - - - - - - - 429.7485
H20BP 373 373 15 373 15 373 15 373 15 373.15 - -
Ga TP - - - - - - - 302.9146
H2 TPe - - - (273 16) 273 16 273.16 - 273.16
H20FP 273 273 15 273 15 - - - - -
Hg TP - - - - - - - 234.3156
2 BPf - 90 18 90 18 90 18 90 188 90.188 - -
Ar TP - - - - - (83.798) - 83.8058
2 TP - - - - 54 361 54.361 - 54.3584
Ne BP - - - - 27 102 27.102 21 102 -
Ne TP - - - - - - 24 5591 24.5561
H2 BP - - - - 20 28 20.28 20 2734 20.3
H2 BP* - - - - 17 042 17.042 17 0373 17.0
H2 TP - - - - 13 81 13.81 13 8044 13.8033
Pb SPh - - - - - - 7 1999 -
*HeBP - - - - - - 4 2221 4.2
In SP - - - - - - 3 4145 -
3HeBP - - - - - - - 3.2
Al SP - - - - - - 1 1796 -
Zn SP - - - - - - 851 -
Cd SP - - - - - - 519 -
a NHS = Normal hydrogen scale [25].b For a time, the ice point was taken to be 273.16 °K. Here, the value 273.15 K
was used to convert temperature values in degrees Centigrade or Celsius to
kelvins in order to be as consistent as possible throughout the table.c FP = Freezing point.d BP = Boiling point at 101,325 Pa.e TP = Triple point.f Changed in 1975 to the condensation point.8 Reduced-pressure boiling point, at P = 33,330.6 Pa.h SP = Superconductive transition point.
1 053 447980 106
676 380
372 692
151 656
002 263
006 596088 966004 770
054 943
7 3
4 3
1 392 408527 153
166 756
050 988
026 514001 975
017 976005 409013 259
5 6
2 9
3 146 6311 357 655
413 923
091 159016 349001 826
-0 004 325-0 004 973
10 3
1 9
Table 4. Values of the coefficients AL and of the constants B and C for the 3He and AHe
vapor -pressure equations and the temperature range for which each equation is valid
Coef. or 3He AHe AHe
Constant 0.65 K to 3.2 K 1.25 K to 2.1768 K 2.1768 K to 5.0 K
A
Ai
A2
A3
A,
A5
A6
A7
A8
A9
B
C
2.2 BETWEEN 3.0 K AND 24.5561 K (THE TRIPLE POINT OF Ne) :
3He and AHe CONSTANTVOLUME GAS THERMOMETRY
Between 3.0 K and 24.5561 K, the ITS -90 is defined in terms of the 3He or AHeconstant volume gas thermometer (CVGT) . The thermometer is calibrated at threetemperatures — at the triple point of neon (24.5561 K) , at the triple point ofequilibrium hydrogen (13.8033 K) , and at a temperature between 3.0 K and 5.0 K,
the value of which is determined by using either a 3He or a AHe vapor-pressurethermometer as described in section 2.1.
For a AHe CVGT that is used between 4.2 K and the triple point of neon(24.5561 K) , r90 is defined by the equation:
r90= a + bp + cp2
, (4)
where p is the CVGT pressure and a, b, and c are coefficients that are determinedfrom calibrations at the three specified temperatures, but with the additionalrequirement that the calibration with the vapor-pressure thermometer be made ata temperature between 4.2 K and 5.0 K.
For a AHe CVGT that is used between 3.0 K and 4.2 K, and for a 3He CVGT that is
used between 3.0 K and 24.5561 K, the non- ideality of the gas must be accountedfor, using the respective second virial coefficients, £ 4 (X90 ) or £ 3 (r90 ) . r90 is
defined in this range by the equation:
a + bp + cp2
T90 = ~ . (5)
1 + BX (T90 )N/V
10
where p is the CVGT pressure; a, b, and c are coefficients that are determinedfrom calibration at the three defining temperatures; Bx (Tm ) refers to B 3 (r90 ) or
B 4 (r90 ); and N/V is the gas density, in moles per cubic meter, in the CVGT bulb.
The values of the second virial coefficients are given by the followingequations
:
for 3He,
B 3 (Tg0 ) /vfmoV1 = [16.69 - 336.98 (T90/K)
-1
+ 91.04 (r90/K)-2
- 13.82 (r90/K)"3
] 1CT 6, (6)
and for *He,
B 4 (r90 )/m3mol _1 = [16.708 - 374.05 (r90/K)
_1- 383.53 (T90/K)"
2
+ 1799.2 (r90/K)"3
- 4033.2 (r90/K)_* + 3252.8 (T90/K)-
5] 10
-6. (7)
The accuracy of realization of Tgo by using a CVGT is dependent upon the CVGTdesign and the gas density used.
2.3 BETWEEN 13.8033 K (THE TRIPLE POINT OF EQUILIBRIUM HYDROGEN) AND 1234.93 K(THE FREEZING POINT OF SILVER) : PLATINUM RESISTANCE THERMOMETRY
Between 13.8033 K (-259.3467 °C) and 1234.93 K (961.78 °C) , the ITS-90 is definedin terms of specified fixed points to which temperature values have beenassigned, by resistance ratios of platinum resistance thermometers obtained bycalibration at specified sets of the fixed points, and by reference functionsand deviation functions of resistance ratios which relate to Tgo between the
fixed points. (Henceforth, for convenience, the standards type platinumresistance thermometers will be referred to by the acronym SPRT.)
2.3.1 GENERAL RELATION BETWEEN RESISTANCE RATIOS AND T90
Temperatures on the ITS-90 in the above -indicated range are expressed in termsof the ratio of the resistance R(Tgo ) at temperature T90 and the resistancei?(273.16 K) at the triple-point temperature of water. (Hereinafter, forconvenience, the terms triple -point temperature, freezing-point temperature andmelting-point temperature will be expressed as triple point, freezing point andmelting point, respectively.) The resistance ratio W(Tgo ) is:
W(Tgo ) = R(Tgo )/R(273.16 K).
The temperature Tgo is calculated from the resistance ratio relation:
W(Tgo ) - WX (TM ) = AW(Tao),
(8)
(9)
where f/(T90 ) is the observed value, Wx (Tgo ) is the value calculated from the
reference functions, and AW(Tgo ) is the deviation of the observed W(TQ0 ) value ofthe particular SPRT from the reference function value. The official version ofthe ITS-90 [83] does not indicate the difference [W(T90 ) - tfr (T90 )] by Mf(TQQ )
.
11
Note that in the earlier international scales, W(T) was defined with referenceto the SPRT resistance at 273.15 K, not 273.16 K.
There are two reference functions Wr (Tgo ) , one for the range 13.8033 K to
273.16 K and another for the range 273.15 K to 1234.93 K. These referencefunctions will be described in the discussion of the two ranges (sees. 2.3.3 and2.3.4).
The deviation AW(Tg0 ) is obtained as a function of T90 for various ranges andsubranges by calibration at specified fixed points. The form of the deviationfunction depends upon the temperature range of calibration.
A schematic representation of the ITS -90 in the range of temperature specifiedfor SPRTs is given in figure 3 . Shown in figure 3 are the temperatures of the
defining fixed points in this region of the scale and the various subrangesspecified by the scale.
2.3.2 SPRT SPECIFICATIONS
The SPRT sensing element must be made from pure platinum and be strain- free.The finished SPRT must meet one of the following criteria:
17(302.9146 K) > 1.118 07, or (10)
17(234.3156 K) < 0.844 235. (11)
These criteria are equivalent to a requirement on the slope, namely,
[aW(r90 )/dr90 ]> 3.986 x 10' 3 K" 1 at 273.16 K. (12)
An SPRT that is acceptable for use to the freezing point of silver must meet thefollowing additional criterion:
17(1234.93 K) > 4.2844. (13)
The temperature range over which an SPRT can be used depends upon the thermometerdesign, but no single design of SPRT can be used over the whole temperature rangewith high accuracy. For measurements at temperatures from 13.8033 K(-259.3467 °C) to 429.7485 K (156.5985 °C) , or perhaps to as high as 505.078 K(231.928 °C) , capsule-type SPRTs that have a nominal resistance of 25.5 at273.16 K may be used. Long-stem type SPRTs of the same nominal resistance maybe used in the range from about 77 K to 933.473 K (660.323 °C) . For temperaturesabove about 660 °C and to as high as 1234.93 K (961.78 °C) , long-stem type SPRTshaving a nominal resistance of 0.25 (or possibly 2.5 0) at 273.16 K should beused. When SPRTs are used at the highest temperatures, leakage currents throughthe insulation supports of the platinum wire become significant and these resultin shunting of the resistor. The effect of this shunting is reduced by usinglow voltages with low resistance SPRTs.
If the sheath of the long- stem type SPRT is borosilicate glass or stainlesssteel, the SPRT should not be used above the upper calibration temperature limitof 420 °C. If the sheath is Inconel, the upper temperature limit becomes about
12
dd Bv 'e6'»6ZL
dd IV 'eZfr886
— dd UZ *ZZ9Z69
dd "S '8Z0809
dd "I '9SPL'6Zb
— dlfll BO '9fr16209
# dlBH'99l-efre3
a>
(0 cc CO"5 DC
CL cc oo aa o.o
a.
^ 0)a *•
O c
•
it
dl JV '890888
dl zO 'fr89Ef9
dJ.3N'L999'frZ
dA 2H-s '802
dA zH"3'ZI.
dl 2H-9 '880881
cd ao occO w4-1 4-)
CO c•H •rH
w Ooooo5-1
*"
10)
X!
C •HH <4-l
4-)
cd W)—1 cCL
a) •Hn 14-1
p5-1
O a)
4-1
Pr)<u M-i
•i-i O-!-l
•H wC) 0)
<1) 5-1
ooo Cfl
34->
cdT-
0) 5-1
W) a)
q &cfl eM <u
o 4->
o> <i)
(A J34J
4->
*•H t3
3 O cO
** <T\(0 Ma> c/i
CD
Q. H W)
E i—
i
1-a)
4->
U
oolt-1
O T3cd
C •Ho M-I•H
CDPcO
4->
co
CO -H3 4-1
1) o cdw•i-l u
(0
S-l CO >H
5-1
>cO
a) oX!o
4-)
V uo
U)<4-i
sCll & 0)
O 5-1o A -Hw W 3
o <ter,
e
req
CO cd i-i
S co
CD o o5-1 6 w3 5-1
M CD CD
•H ,C rCta 4-> 4->
13
660 °C. If the sheath is fused silica, temperature measurements can be made up
to 962 °C.
For measurements up to about 630 °C, mica is just barely adequate as anelectrical insulator for the temperature sensing element and leads of SPRTs.
Starting at about 500 °C, muscovite mica begins to decompose, evolving water that
electrically shunts the thermometer resistor. Phlogopite mica is adequatelystable to 630 °C. For measurements up to 962 °C, refractory materials such as
fused silica and sapphire are used for electrical insulation.
2.3.3 RANGE 13.8033 K TO 273.16 K
In the range 13.8033 K to 273.16 K, the reference function Wr (Tgo ) is given by:
12
£n[Wr (TQ0 )] = A + I AiUinCTgo/273.16 K) + 1.5J/1.5}1
. (14)i=l
A specified, approximate inverse [83] of this equation, equivalent to within± 0.000 1 K, is:
15
T90/273.16 K = B + I BiUt^CTgo)] 1/6- . 65)/0 . 35} 1
. (15)i=l
The values of the constants A and B , and of the coefficients AL and EL ofequations (14) and (15) are listed in table 5.
If an SPRT is to be used throughout the range from 13.8033 K to 273.16 K, it mustbe calibrated at the triple points of equilibrium hydrogen (13.8033 K) , neon(24.5561 K) , oxygen (54.3584 K) , argon (83.8058 K) , mercury (234.3156 K) , andwater (273.16 K) , and at two additional temperatures close to 17.0 K and 20.3 K.
The temperatures of calibration near 17.0 K and 20.3 K may be determined by usingeither a CVGT as defined in section 2.2 or the specified vapor-pressure-temperature relation of equilibrium hydrogen.
When the CVGT is used, the two temperatures must be within the ranges 16.9 K to
17.1 K and 20.2 K to 20.4 K, respectively. When the equilibrium hydrogen vapor-pressure thermometer is used, the two temperatures must be within the ranges17.025 K to 17.045 K and 20.26 K to 20.28 K, respectively. The temperatures ofthe equilibrium hydrogen vapor-pressure thermometer are determined from the
values of the hydrogen vapor pressure, p, and the equations:
T90/K - 17.035 = (p/kPa - 33 . 3213)/13. 32 (16)
r90/K - 20.27 = (p/kPa - 101.292)/30, (17)
where 13.32 and 30 are values of (dp/dT90)/(kPa/K) at 17.035 K and 20.27 K,
respectively.
Depending upon the temperature range of use, the SPRT may be calibrated from273.16 K down to 13.8033 K (the triple point of equilibrium hydrogen), down to
14
Table 5. Values of the coefficients Ait Ei
Ci and Di and of the constants Ao»
Bo> '0' and D in the reference functions, eqs (14) and (22), and in the
functions approximating them, given by eqs (15) and (23)
Constant orCoefficient
Value Constant orCoefficient
Value
2.135 347 29
3.183 247 20
1.801 435 97
0.717 272 04
B r 0.183 324 7220.240 975 303
0.209 108 7710.190 439 972
^5A6
A7
An
H2
0.503 440 27
-0.618 993 95
-0.053 323 22
0.280 213 62
0.107 152 24-0.293 028 65
0.044 598 72
0.118 686 32
-0.052 481 34
2.781 572 54
1.646 509 16
•0.137 143 90
0.006 497 67
'10
»12
»15
0.142 648 4980.077 993 4650.012 475 611-0.032 267 127
-0.075 291 522-0.056 470 6700.076 201 285
0.123 893 204
-0.029 201 193-0.091 173 542
0.001 317 6960.026 025 526
439.932 854472.418 02037.684 4947.472 018
0.002 344 440.005 118 68
0.001 879 82
0.002 044 72
2.920 828
0.005 1840.963 8640.188 732
0.000 461 22
0.000 457 24
0.191 203
0.049 025
24.5561 K (the triple point of neon), down to 54.3584 K (the triple point ofoxygen), or down to 83.8058 K (the triple point of argon).
The deviation function for calibration over the range 13.8033 K to 273.16 K is
given by the relation:
A^(r90 ) = ai [tf(r90 ) - l] + b1[t^(r90 ) - l]
2 + I Ci [im/(r90 )]i+n
,
i=l(18)
15
with n = 2. The coefficients a1( blf and the five c i' s are obtained by
calibration at all eight of the above temperatures, including that at the triplepoint of water. The values of Wr (TQ0 ) are obtained from the reference function[eq (14)].
The official version of the ITS-90 [83] does not indicate the difference[W(TQ0 ) - Wx (Tgo )] by AWm (TQQ ) , does not use the subscript m, where in eq (18),
m = 1, nor does it label the coefficients a and b with subscript m. However,we adopt this subscript notation to identify the subranges. Additionally, thisnotation is useful when reporting calibration results.
If an SPRT is not to be used over the entire 13.8033 K to 273.16 K range, butits use restricted to one of the subranges, the deviation functions and thecalibration points are as follows.
2.3.3.1 SUBRANGE 24.5561 K TO 273.16 K
The deviation function for calibration in the subrange 24.5561 K to 273.16 K is
given by the relation:
3
Atf2 (T90 ) = a2 [W(T9Q ) - 1] + b2 [W(T9Q ) - l]2 + £ c± [£vW(Tgo ) }
i+n, (19)
i=l
with n = 0. The coefficients a2 , b2 , and the three c^s are obtained bycalibrating the SPRT at the triple points of equilibrium hydrogen (13.8033 K)
,
neon (24.5561 K) , oxygen (54.3584 K) , argon (83.8058 K) , mercury (234.3156 K)
and water (273.16 K) . The values of tfr (r90 ) are obtained from the referencefunction [eq (14)]. Note that for this subrange, temperatures are measured onlydown to the triple point of neon, although the thermometer must be calibratedat the triple point of equilibrium hydrogen.
2.3.3.2 SUBRANGE 54.3584 K TO 273.16 K
The deviation function for calibration in the subrange 54.3584 K to 273.16 K is
given by the relation:
W3 (Tgo ) = a3 [W(TQ0 ) - 1] + b 3 [W(T30 ) - l] 2 + Cl [intf(r90 )]1+n
, (20)
with n = 1. The coefficients a3 , b 3 , and c± are obtained by calibrating the SPRTat the triple points of oxygen (54.3584 K) , argon (83.8058 K) , mercury(234.3156 K) , and water (273.16 K) . The values of Wr (T30 ) are obtained from thereference function [eq (14)].
2.3.3.3 SUBRANGE 83.8058 K TO 273.16 K
The deviation function for calibration in the subrange 83.8058 K to 273.16 K is
given by the relation:
Atf4 (T90 ) = ajtf(r90 ) - 1] + b A [tf(r90 ) - l]intf(r90 ). (21)
The coefficients a 4and b
4are obtained by calibrating the SPRT at the triple
16
points of argon (83.8058 K) , mercury (234.3156 K) , and water (273.16 K) . The
values of f/r (T90 ) are obtained from the reference function [eq (14)].
2.3.4 RANGE 273.15 K (0 °C) TO 1234.93 K (961.78 °C)
In the range 273.15 K to 1234.93 K, the equation for the reference functionWr (Tgo ) is given by:
9
Wr (T90 ) = C + I CjCTgo/K - 754.15)/481] i. (22)
i=l
A specified, approximate inverse [83] of this equation, equivalent to within± 0.000 13 K, is:
9
T90/K - 273.15 = D + I DiU^CTgo) - 2 . 64]/l . 64} 1. (23)
i=l
The values of the constants C and D and of the coefficients C tand D ± of eqs
(22) and (23) are listed in table 5.
If the SPRT is to be used over the entire range 273.15 K to 1234.93 K, it mustbe calibrated at the triple point of water (273.16 K) and at the freezing pointsof tin (505.078 K) , zinc (692.677 K) , aluminum (933.473 K) , and silver(1234.93 K).
The deviation function is given by the relation:
W6 (T90 ) = a6 [tf(r90 ) - 1] + b 6 [tf(T90 ) - l]2
+ c 6 [W(T90 ) - l]3 + d[W(Tao ) - ^(933.473 K)] 2
. (24)
The values of a6 , b 6 , and c 6 are determined from the measured deviations &W(T90 )
of W(T30 ) from the reference values Wr (Tgo ) [obtained from eq (22)] at the
freezing points of tin (505.078 K) , zinc (692.677 K) , and aluminum (933.473 K)
,
ignoring the term involving d. Then, d is determined from these values of a6 ,
b6 , and c6 and the deviation AP/(r90 ) of W(TgQ ) from the reference value tfr (T90 ) at
the freezing point of silver (1234.93 K) . The coefficient d is used only fortemperature measurements in the range from the freezing point of aluminum to the
freezing point of silver. For temperature measurements below the freezing pointof aluminum, d = 0.
SPRTs may be calibrated for use throughout the whole range 273.15 K to 1234.93 Kor for shorter subranges by calibrations at fixed points between 273.15 K andthe upper limit of 933.473 K (freezing point of aluminum, 660.323 °C) , of692.677 K (freezing point of zinc, 419.527 °C) , of 505.078 K (freezing point oftin, 231.928 °C) , of 429.7485 K (freezing point of indium, 156.5985 °C), or of302.9146 K (melting point of gallium, 29.7646 °C)
.
The deviation function Afof5 (r90 ) will be discussed later in the text.
17
If an SPRT is not to be used over the entire 273.15 K to 1234.93 K range, butits use restricted to one of the subranges, the deviation functions and the
calibration points are as follows.
2.3.4.1 SUBRANGE 273.15 K (0 °C) TO 933.473 K (660.323 °C. FREEZING POINT OFALUMINUM)
For an SPRT used throughout the subrange 273.15 K to 933.473 K, the thermometeris calibrated at the triple point of water (273.16 K) , and at the freezing pointsof tin (505.078 K) , zinc (692.677 K) , and aluminum (933.473 K) . The deviationfunction is given by the relation:
AW7 (Tgo ) = a7[W(Tgo ) - 1] + b 7 [W(Tgo ) - l] 2 + c 7 [W(Tgo ) - l] 3
. (25)
The coefficients a7 , b 7 , and c 7 , identical to a6 , b6 and c 6 , respectively, are
determined from the deviations AW(Tgo ) of W(Tgo ) from the reference values Wr (Tgo )
[eq (22)] at the freezing points of tin (505.078 K) , zinc (692.677 K) , andaluminum (933.473 K)
.
2.3.4.2 SUBRANGE 273.15 K (0 °C) TO 692.677 K (419.527 °C. FREEZING POINT OFZINC)
For an SPRT used throughout the subrange 273.15 K to 692.677 K, the thermometeris calibrated at the triple point of water (273.16 K) , and at the freezing pointsof tin (505.078 K) and zinc (692.677 K) . The deviation function is given by therelation:
AfcT8 (r90 ) = a8 [f/(r90 ) - 1] + ba [W(Tgo ) - l] 2. (26)
The coefficients a8 and b 8 are determined from the deviations AW(Tgo ) of W(Tgo )
from the reference values Wr (Tgo ) [eq (22)] at the freezing points of tin(505.078 K) and zinc (692.677 K)
.
2.3.4.3 SUBRANGE 273.15 K (0 °C) TO 505.078 K (231.928 °C. FREEZING POINT OFTIN)
For an SPRT used throughout the subrange 273.15 K to 505.078 K. the thermometeris calibrated at the triple point of water (273.16 K) , and at the freezing pointsof indium (429.7485 K) and tin (505.078 K) . The form of the deviation functionis the same as that for the subrange 273.15 K to 692.677 K, i.e.,
M/9 (r90 ) = a9 [tf(r90 ) - l] + b 9 [tf(r90 ) - i] 2. (27)
The coefficients a9 and b 9 are determined from the deviations AW(Tgo ) of W(Tgo )
from the reference values Wr (Tgo ) [eq (22)] at the freezing points of indium(429.7485 K) and tin (505.078 K)
.
2.3.4.4 SUBRANGE 273.15 K (0 °C) TO 429.7485 K (156.5985 °C. FREEZING POINT OFINDIUM)
For an SPRT used throughout the subrange 273.15 K to 429.7485 K, the thermometeris calibrated at the triple point of water (273.16 K) and at the freezing point
18
of indium (429.7485 K) . The deviation function is:
W10 (T90 ) = a10 [tf(T90 ) - 1]. (28)
The coefficient a10 is determined from the deviation Af/(r90 ) of W(Tgo ) from the
reference value Wr (Tg Q ) [eq (22)] at the freezing point of indium (429.7485 K) .
2.3.4.5 SUBRANGE 273.15 K CO °C) TO 302.9146 K (29.7646 °C. MELTING POINT OFGALLIUM)
For an SPRT used throughout the subrange 273.15 K to 302.9146 K, the thermometeris calibrated at the triple point of water (273.16 K) and at the melting pointof gallium (302.9146 K) . The deviation function is:
Atfii(r90 ) = an [^(T90 ) - 1]. (29)
The coefficient an is determined from the deviation A^(T90 ) of W(Tgo ) from the
reference value Wr (TgQ ) [eq (22)] at the melting point of gallium (302.9146 K) .
2.3.5 SUBRANGE 234.3156 K (-38.8344 °C, TRIPLE POINT OF MERCURY) TO 302.9146 K(29.7646 °C. THE MELTING POINT OF GALLIUM)
For an SPRT used throughout the subrange 234.3156 K to 302.9146 K, thethermometer is calibrated at the triple points of mercury (234.3156 K) and water(273.16 K) , and at the melting point of gallium (302.9146 K) . The form of thedeviation function is the same as that for the subrange 273.15 K to 692.677 K,
i.e.,
Atf5 (r90 ) - a5 [W(T9Q ) - 1] + b 5 [f7(r90 ) - l]2
. (30)
The coefficients a5 and b 5 are determined from the deviations AJ7(r90 ) of W(TQ0 )
from the reference values WZ (TQ0 ) at the triple point of mercury (234.3156 K) andat the melting point of gallium (302.9146 K) . The reference values WT (T9Q ) mustbe calculated from the appropriate reference function [either eq (14) or eq
(22)], both reference functions being required to cover this range.
2.4 ABOVE 1234.93 K (961.78 °C. FREEZING POINT OF SILVER): RADIATION THERMOMETRYBASED ON PLANCK'S LAW OF RADIATION
At temperatures above 1234.93 K, r90 is defined by the relation:
LA (r90 ) exp[c2/AT90 (X)] - 1
LA [r90 (X)] exp[c2/Ar90 ]- 1
(31)
in which LX (TQ0 ) and LA [r90 (X)] are the spectral concentrations of the radianceof a blackbody at wavelength A (in vacuum) at TQ0 and at T90 (X), respectively.Tgo (X) refers to either the silver freezing point [r90 (Ag) = 1234.93 K] , the goldfreezing point [T90 (Au) - 1337.33 K] or the copper freezing point[T90 (Cu) = 1357.77 K] . The second radiation constant, c2 (= hc/k) , of Planck'sradiation formula has the value c2 = 0.014388 m»K. Although the freezing-point
19
temperature of silver is the junction point of platinum resistance thermometryand radiation thermometry, it is believed that the rgo values of the freezingpoints of silver, gold and copper are sufficiently self-consistent that the useof any one of them as the reference temperature T90 (X) will not result in anysignificant difference in the measured values of T90 from what would be obtainedif only the silver freezing point were used.
3. REALIZATION OF THE ITS -90
3.1 VAPOR PRESSURE THERMOMETRY AND GAS THERMOMETRY
3.1.1 REALIZATION OF THE ITS -90 BELOW 273.16 K
The calibration of thermometers below the triple point of argon on the ITS -90,
as defined, is relatively complex. It is expected that capsule-type SPRTs
,
rhodium-iron resistance thermometers (RIRTs) , and other stable encapsulatedresistance thermometers will be calibrated in terms of the defined ITS -90 andthen used as reference thermometers to maintain the ITS -90 below about 84 K andused to calibrate other resistance thermometers by the comparison method. Thereference thermometers will be calibrated periodically in terms of the definedITS -90.
By use of the term "realization of the ITS-90," reference is made to obtainingthe equilibrium states as defined by the scale, to having thermometers in thermalequilibrium with those equilibrium states, and to making accurate measurementsand interpretations of the requisite properties of those thermometers in termsof the ITS-90.
Considerable effort has been expended to develop and realize the EPT-76, a scalewhich covered the range 0.5 K to 30 K. This scale has been widely disseminatedamong low temperature scientists. At NIST, the EPT-76 has been maintained onreference -standard RIRTs for use in calibrating customer thermometers. Uponintroduction of the ITS-90, NIST converted the EPT-76 on the reference -standardRIRTs to the ITS-90 using the specified differences [83] between Tgo and T76 .
This converted scale is being used to calibrate other thermometers until suchtime that NIST realizes the ITS-90 in this temperature region directly asdefined. It is recommended that those laboratories that have thermometers withcalibrations on the EPT-76 adjust their T76 values to conform to r90 values.When NIST realizes the ITS-90 as defined in this range, the difference betweenthe converted scale on the reference-standard RIRTs (and where appropriate, oncapsule SPRTs) and the realized scale will be determined.
3.1.2 VAPOR PRESSURE THERMOMETRY AND THE CVGT RANGE
For most measurements below about 100 K, better precision can be obtained withcapsule- type SPRTs than with the long- stem type. In the calibration of SPRTs,
however, long-stem type SPRTs (immersion- type SPRTs) can be calibrated easilyby a direct immersion process down to the triple point of argon (83.8058 K) bymoving the SPRTs from one fixed-point device to another. Unless capsule -typeSPRTs and other capsule -type thermometers are installed inside long stem- like
holders, however, they will require re- installation and re-wiring wheneverdifferent fixed-point apparatuses are used. (In this document, the phrase
20
"capsule-type thermometers" means encapsulated resistance thermometers of small
overall dimensions.) It is desirable, therefore, to be able to calibratecapsule-type SPRTs and other capsule-type thermometers at the argon triple pointand below (or, if possible, even at the triple point of water and below sincecalibrations of SPRTs require measurements at the triple point of water) in anintegrated "multi-task" (multi- fixed-point) apparatus. Such a multi-taskapparatus requires, in addition to wells for capsule thermometers (in thermalequilibrium in a "single cryostat block"), means for calibration using 3He, AHe
,
and e-H2 vapor pressure thermometry, 3He and *He CVGTs , and triple points of e-H2 ,
Ne, 2 , Ar, Hg, and H20. A total of 11 chambers is required if all of the
overlapping definitions of the ITS -90 are to be evaluated and if a "continuouscalibration," without re- installation and re-wiring of the SPRTs, is desired from0.65 K to 273.16 K. In addition, unless high pressure sealed cells of the
reference gases are used, tubes to each of the chambers, except those for Hg andH20, are required.
Since 3He and 4He vapor-pressure and CVGT ranges overlap to a large extent,chambers for AHe vapor-pressure measurements and for 4He CVGT measurements couldbe eliminated and the ITS -90 could still be realized. Also, chambers for e-H2
vapor-pressure and the e-H2 triple-point realizations could be combined. Thenumber of chambers could be reduced further if the cryostat block could beallowed to warm to ambient temperature or higher for exchanging certain of the
fixed-point substances, or if it were permissible to perform the calibrationsof the capsule thermometers at the triple points of argon, mercury, and waterin other apparatuses, using a long- stem type holder. This procedure, however,would require a longer time for calibration.
The number of chambers can also be reduced if suitable, highly stable capsulethermometers are available for correlating the scales; for this purpose, thecapsule SPRTs would be calibrated in another, simpler, fixed-point apparatus.It is expected that such a set of reference -standard resistance thermometerswould be calibrated, and that routine calibrations of customer thermometers onthe ITS -90 would be by comparison with these reference thermometers. Thereference thermometers would be checked occasionally against the defined ITS -90.
It is hoped that resistance thermometer devices will be more reproducible thanthe ability to realize the defined ITS -90. Depending upon the design of thecryostat, the defined ITS -90 may lack the desired reproducibility when realizedin a multi -task apparatus of high complexity. In order to achieve the bestrealization of the ITS -90, it may be more practical to limit the number ofdefining fixed points in a single cryostat block.
The greatest problem in realization of the fixed points and in calibrations ofcapsule thermometers is ensuring that the multi-task or a "single-task" cryostatblock is isothermal. Depending upon the design, a thermal gradient can persist.The presence of 3He, AHe , e-H2 , Ne , 2 , and/or Ar gases in their respectivechambers are expected to be beneficial in making the apparatus isothermal, but,the gases may be a source of thermal oscillation. (In designing an apparatusfor low temperature gases, provisions should be made to avoid thermaloscillations in the gas.) The vapor pressures of Ne and Ar are high at theirrespective triple-point temperatures so that thermal equilibrium should be easilyattained at their triple points.
21
For a practical cryostat block, the 3He and 4He CVGTs must connect thermally the3He and 4He vapor-pressure scales and the fixed points of e-H2 and Ne . Such a
cryostat block will require six chambers (separate 3He and AHe vapor-pressurechambers, separate 3He and 4He CVGT chambers, e-H2 vapor-pressure and e-H2 triple-
point chamber, and a Ne chamber) to realize the ITS -90 in the most satisfactorymanner at and below the Ne triple point. This arrangement will permit a directcomparison of the different parts of the scale and/or calibration ofthermometers. Since the combination of 4He vapor-pressure thermometry and 4Heconstant volume gas thermometry are redundant with 3He vapor-pressure thermometryand 3He constant volume gas thermometry, the 4He systems are not required.Hence, the number of chambers required could be reduced to four and the ITS -90
could still be realized. It should be noted, however, that, depending upon the
CVGT filling pressure, the dp/dT of a 3He CVGT may be less sensitive than *He
vapor-pressure thermometry. Since SPRTs can be calibrated only down to the argontriple-point temperature using the long-stem SPRT apparatus, it would be mostpractical and useful to include an oxygen triple -point chamber in the low-
temperature system, thereby increasing the number of chambers to five. Also,since it is highly desirable to overlap calibrations obtained in a long- stem typeSPRT apparatus with those obtained in a low- temperature apparatus, an argontriple-point chamber should be included. This increases the number of chambersfor the low- temperature apparatus to six.
Although it is desirable to have as few tubes as possible going to the cryostatblock in order to minimize temperature gradients in the block, one must buildinto the system enough redundant components to be able to check the system forproper and accurate operation. For example, although the CVGT is calibrated atonly the triple -point temperatures of neon and hydrogen and at one point in the3He or ''He vapor pressure range between 3.0 K and 5.0 K, the system should havethe capability for the measurement of hydrogen vapor pressures at about 17 K and20.3 K so that temperatures measured by means of vapor pressures and by the CVGTmay be compared. Of course, if the system is operating properly, thetemperatures measured by the two techniques should agree. Similarly, thereshould be the capability to measure the vapor pressures of both 3He and AHe so
that temperatures measured with the CVGT in the range from 3 K to 5 K and bymeans of 3He and AHe vapor pressures may be compared for agreement. Also, it is
desirable to design the CVGT for absolute gas thermometry measurements; the CVGTcan check the consistency of the ITS -90 from about 3 K to 90 K.
3.1.3 REALIZATION OF THE VAPOR PRESSURE AND CVGT SCALES AT TEMPERATURES BELOWTHE NEON TRIPLE POINT
Since measurement of pressure is common to both vapor pressure and CVGTmeasurements, it will be discussed later in this section (see sec. 3.1.3.5).
In vapor pressure measurements, it is important that cold spots be absent alongthe gas -pressure transmitting tube. If cold spots are present, the observedvapor pressure will reflect the temperature of the condensation at the cold spotinstead of that of the bulk bath. A separate vacuum jacket around the tube willmaintain a continuous heat flux to the sample bulb or bath and should free the
tube of any condensation [24]. The vacuum jacket should also reduce theoccurrence of thermal oscillation in the gas -pressure sensing tube. If thermaloscillations do occur, they may be suppressed by either one or a combination of
22
the following: increasing the volume of the external gas -pressure space at the
ambient temperature, or by inserting a wad of wool or glass fiber or a piece of
yarn in the gas -pressure sensing tube. The thermal oscillations may be
suppressed also by "tuning" a variable volume device [36] . Thermocouples shouldbe placed along the gas -pressure sensing tube in order to determine temperaturesalong that tube, the distribution of those temperatures being required to
determine the aerostatic head correction.
The vapor pressure may be determined over the bath of bulk liquid 3He ,
AHe, or
e-H2 with which the thermometer to be calibrated is in thermal equilibrium. Themeasurement can be made also by using a separate, small sample bulb with whichthe thermometer is in good thermal contact. The latter method is preferred withthe rather expensive 3He and with e-H2 , which requires a catalyst for the
equilibrium ortho-para conversion of the sample [2,58,91].
3.1.3.1 3He Vapor- Pressure Measurements
Because of the relatively high cost of the sample, vapor-pressure measurementsare made with 3He contained in a small volume of about 5 cm3
. Likewise, the gas-
pressure volume to the manometer should be kept relatively small, but largeenough to avoid large thermomolecular pressure effects and thermal oscillations.Thermomolecular-pressure-effect corrections depend on the sensing tube diameter,surface condition of the tube, temperature difference, and pressure [49,71,102].As mentioned above, thermal oscillations can be reduced by varying the gas-pressure volume at the ambient temperature or by introducing a wad of fiber or
yarn (cotton, wool, or glass) in the gas-pressure tube.
In the past, 3He contained significant amounts of AHe and the observed vaporpressures of 3He required corrections for its presence. In recent years,however, 3He samples of 99.9995% purity have become available, eliminating the
requirement for such corrections. At 0.65 K, the vapor pressure and the
temperature derivative of the vapor pressure are 115.9 Pa and 1.08 Pa/mK,
respectively. At the upper limit of 3.2 K, the vapor pressure and the
temperature derivative of the vapor pressure are 101,662.1 Pa and 106.83 Pa/mK,respectively. The required pressure resolution that corresponds to 0.1 mK ofvapor-pressure measurements varies from 0.108 Pa at 0.65 K to 10.7 Pa at 3.2 K.
Since the amount of 3He in the cryostat is small, and since the amount of AHeused in cooling is relatively large, every effort should be made to avoidcontamination of the sample of 3He by the AHe through diffusion, particularlythrough any glass parts of the apparatus.
The sample bulb should contain enough 3He that the liquid surface temperature andthe cryostat block temperature can be correlated with the observed vaporpressure. The temperature of the cryostat block must be checked for consistencywith the observed dp/dT of the vapor pressure at the temperature of measurement.
Aerostatic-head corrections depend upon the density of the gas in the pressure-transmitting gas tube. Thermocouples must be distributed along the tube in orderto measure the temperatures required for calculation of these corrections.
23
3.1.3.2 He Vapor -Pressure Measurements
Since liquid 4He can be obtained easily, the vapor pressure can be determinedabove a bath of the liquid in which an apparatus containing the thermometer is
immersed. Or, a technique using a small bulb of 4He sample, with thethermometer in good thermal contact, can be employed. The lower end of the gas-
pressure tube should have a small orifice in order to reduce superfluid 4He filmflow at temperatures below 2.1768 K [61,91]. The AHe sample bulb must be inthermal equilibrium with the cryostat block. The cryostat block temperatureshould be checked for consistency with the observed dp/dT of the vapor pressureat the temperature of measurement.
At the lower temperature limit of 1.25 K, the vapor pressure of AHe is 114.7 Paand the temperature derivative of the vapor pressure is 0.76 Pa/mK. At theupper limit of 5 K, the vapor pressure and the temperature derivative of the
vapor pressure of 4He are 194629.7 Pa and 146.53 Pa/mK, respectively. Therequired pressure resolutions of the vapor pressure that corresponds to 0.1 mKare 0.076 Pa at 1.25 K and 14.7 Pa at 5.0 K [37,38].
3.1.3.3 e-H2 Vapor Pressure Measurements
The equilibrium composition of the two molecular states of hydrogen (ortho andpara) is temperature dependent. The room temperature composition, about 75%
ortho and 25% para, is referred to as normal hydrogen (n-H2 ) . On liquefaction,the composition slowly changes toward the equilibrium composition correspondingto its temperature. In the process, the heat of transition is released,resulting in the evaporation of some hydrogen. A catalyst, such as activatedferric hydroxide, hastens the equilibration. The catalyst must be placed in the
sample chamber in order to ensure that the hydrogen has the appropriateequilibrium composition. Most of the conversion must be made before collectingthe liquid in the sample chamber since the heat of conversion (1064 J/mol) fromthe ortho to the para molecular state is larger than the heat of vaporization(900 J/mol) of normal hydrogen. The normal boiling point of e-H2 (equilibriumcomposition: 0.21% ortho and 99.79% para) is about 0.12 K lower than that ofn-H2 . The temperatures near 17.035 K and 20.27 K are determined from vapor-pressure measurements near 33,321.3 Pa and 101,292 Pa, respectively [2,31,58].
3.1.3.4 Constant Volume Gas Thermometry
Some of the following precautions and corrections that are applicable to
absolute constant -volume gas thermometry should be included in the calibrationof the CVGT at the three specified temperatures of calibration:
1. The volume of the gas bulb should be sufficiently large relative to the gas-pressure-line volume to minimize the error in correcting for the "dead space."On the other hand, the diameter of the gas -pressure line should not be so smallas to cause large thermomolecular pressure corrections
.
2. The temperature coefficient of volume expansion and the pressure expansionof the gas bulb should be known accurately. (It is desirable to check thecalibration by using the CVGT in the absolute mode.)
24
Table 6. The effect of pressure on the temperatures of the defining fixedpoints. The reference pressure for the equilibrium states of freezing andmelting points is one standard atmosphere (101,325 Pa). Triple points have the
vapor pressure of the material when the solid, liquid and vapor phases are
present in equilibrium
Material 90
e-H2 TP
Ne TP
o2 TP
Ar TP
Hg TP
H2 TP
Ga MP
In FP
Sn FP
Zn FP
Ai FP
Ag FP
Au FP
Cu FP
13.
24,
54.
83.
234.
273.
302.
429,
505.
692.
933.
1234.
1337.
1357.
8033
5561
3584
8058
3156
16
9146
7485
078
677
473
93
33
77
Pressure Effect of Fixed Point
K Pa" 1 mK/(meter of liquid)xlO8*
34
16
12
25
5.
-7.
2.04.9
3.3
4.3
7.0
6.0
6.1
3.3
0.25
1.9
1.5
3.3
7.1
-0.73
-1.2
3.3
2.2
2.7
1.6
5.4
10.
2.6
* Equivalent to millikelvins per standard atmosphere
3. In order to be able to calculate the aerostatic head correction, the
temperature distribution along the connecting gas -pressure transmitting tube(capillary) must be known. That temperature distribution may be determined byplacing thermocouples along the tube.
4. The gas -bulb filling pressure should be sufficiently high to give anadequate dp/dT for measurement, but not so high as to require large correctionsfor non- ideality of the gas.
5. Higher pressures reduce the thermomolecular pressure gradients in theconnecting gas -pressure tube.
6. The effect of adsorption can be reduced by designing the gas bulb so thatthe volume is large relative to the surface and by polishing the inside surfaceof the bulb
.
For optimizing the CVGT design, the ideal gas law may be applied.Differentiating the ideal gas relation,
25
pV = riRT, (32)
yields
dp/dT = Rn/y, (33)
where p is the pressure of n moles of gas contained in a volume of V m3.
Equation (33) shows that the dp/dT sensitivity of the CVGT is directly relatedto the gas density n/V. Expressing the gas constant R as 8.31441 Nm/mol»K or8.31441 Nm3/(m2 »mol»K) , the sensitivity dp/dT can be expressed in the units Pa/K.
Thus,
dp/dT - 8.31441(n/7) Pa/K, (34)
where n/V is given in mol/m3. If a gas bulb of 1000 cm3 is filled to four
atmospheres at 273.16 K, n/V would be approximately 178 mol/m3, and dp/dT
becomes 1484 Pa/K. Since the resolution of many high quality mercury-manometersystems is about 0.03 Pa to 0.2 Pa, the temperature resolution is about 0.03 mKto 0.1 mK. It is to be noted that since
p/T - Rn/V, (35)
the sensitivity dp/dT is "constant," independent of the gas bulb volume, as longas the gas density at the filling temperature is constant. Hence, the gas -bulb
volume and the gas filling pressure should be chosen so that errors from the
effect of dead space, gas non- ideality effects, and other effects are negligiblysmall. See references [6,13,49,60].
3.1.3.5 Pressure Measurements
Efforts have been made to minimize the requirement of highly-precise andaccurate pressure measurements in the realization of the defining fixed pointsof the ITS -90. The fixed points involving freezing and melting requireknowledge of the pressure only within the significance of the relatively smallpressure effect (cf. table 6). Accurate pressure measurements are required,however, to realize the vapor -pressure -temperature scales of 3He and 4He in the
range 0.65 K to 5.0 K, and to realize the vapor-pressure -temperature scale ofe-H2 close to 17.035 K and 20.27 K. To realize the CVGT scale using 3He or AHegas in the range 3.0 K to the triple point of neon (24.5561 K) , only accuratepressure -ratio measurements are required.
3.1.3.5.1 Mercury Manometry
Mercury manometry requires precise determination of the difference in height ofthe two mercury surfaces in a U-tube manometer. Traditionally, cathetometershave been used with a smallest imprecision of about 2 Pa. In recent years, the
levels have been sensed, in conjunction with length standards, by capacitive andinterferometric methods. The resolution of such mercury manometry systems is
about 0.05 Pa [19,50,52,81]. (Note: the NIST manometry resolution has beenreported [50] to be about 0.0013 Pa.) For accurate pressure measurements, it
is necessary to know the density of mercury (which is pressure and temperature
26
dependent), the capillary depression at the mercury meniscus, the vapor pressureof mercury, the aerostatic head difference of the pressure transmitting gas or
gases, and the local acceleration due to gravity. At one standard atmosphere,uncertainties of absolute pressure measurements of about 3 ppm and pressureratios of about 1 ppm have been reported. See references [19,50,52,81].
3.1.3.5.2 Oil Manometry
The techniques and the requirements of oil manometry are similar to those ofmercury manometry.
3.1.3.5.3 Piston Gauges (Pressure Balances)
The pressure balanced by a dead-weight piston gauge is obtained from the massof the piston and the applied weights, and the effective area of the freelyrotating piston inside a closely- fitting cylinder. For determination of the
absolute pressure, the gauge must be enclosed and evacuated by a high capacitypumping system to minimize the back pressure from the gas leaking between thepiston and the cylinder [13,60,81,96]. The local acceleration due to gravitymust be known accurately. Corrections must be applied for the effect of thestreaming gas and for any back pressure. It is advisable to check the readingsof the piston gauge against a primary mercury manometer. Also, the variationof the effective area with pressure must be determined with a mercury manometer.The aerostatic head of the manometry system will change as gas leaks between the
piston and the cylinder causing the piston to sink deeper into the cylinder.The position of the piston may be maintained by automatically pumping more gasinto the system. A resolution of 1 ppm [13] and an uncertainty of about 15 ppmhave been reported in the pressure range 2 kPa to 200 kPa [63,81].
3.1.3.5.4 Diaphragm Pressure Detector
The diaphragm pressure detector consists of a thin metal disk clamped undertension between two flat electrodes which form two capacitors, with the diskcommon to both capacitors. Any pressure differential across the metal diskcauses the disk to deflect, increasing the capacitance on one side whiledecreasing the capacitance on the other side. This change is detected bycapacitance bridge techniques. Instruments for absolute pressure measurementsare available; however, they require periodic recalibrations to achieveuncertainties of 1 to 5 parts in 10* of the readings.
The diaphragm pressure detector is used in high precision manometry as pressurebalance detectors, i.e., with the pressures equal on both sides of thediaphragm. The diaphragm pressure balance detector separates the gas of theapparatus (vapor pressure apparatus or GVGT) from the counter-balancing gas ofwhich the pressure is determined.
The resolution of diaphragm gauges has been reported [13] to be about 0.002 Pa.
Instability due to different pressures, hysteresis, temperature effects, andother causes may decrease the resolution to 0.02 Pa in actual pressuremeasurements [13].
27
3.1.3.5.5 Thermomolecular Pressure Difference
Thermomolecular pressure differences occur at low gas pressures in tubes withtemperature gradients when the tube diameter is not much larger than the meanfree path of the gas . The pressure difference depends upon the gas , the
temperature of the gas , the diameter of the tube , the tube material , and the
surface condition of the tube. The best procedure is either to use a
sufficiently large tube to minimize the thermomolecular pressure difference or
to experimentally determine the difference by comparing the pressures betweenthe small diameter tube being used in the cryostat and a large diameter tube
[49,71,102].
3.2 REALIZATION OF THE FIXED POINTS OF THE ITS -90
3.2.1 EFFECT OF IMPURITIES
Except for the vapor -pressure -temperature points of helium and equilibriumhydrogen, the fixed points of the ITS -90 are freezing points, melting points,or triple points. The vapor-pressure measurements with 3He, 4He, and e-H2 mustbe performed with sufficiently pure samples to minimize the effects ofimpurities. The principal components of air impurity would be frozen. Neon inhydrogen, however, causes positive deviations from ideal behavior [93,97]. Inthe vapor pressure measurements of AHe, it is very likely that the 4He will bepure but 3He may contain some *He. For such a circumstance, Roberts, Sherman andSydoriak described a procedure for correcting for the presence of 4He in 3He
[90].
The temperatures of freezing points (liquid-solid or liquid- solid-vaporequilibrium points) of substances are usually lowered by the presence of animpurity. This sometimes, however, is not the case when that impurity is
soluble in both the liquid and the solid phases of the major component. If a
given impurity is known to be present or is suspected, one must consult the
literature on the heterogeneous phase data of metal and non-metal systems to
estimate the possible effect of that impurity on the freezing point [39,51,88].(Note: often in the analysis of the effect of impurities on freezing points,the impurity is assumed to be nonvolatile.)
Assuming that the ideal solution law holds and that the impurities remain inliquid solution, with no concentration gradients, then as the major componentslowly freezes, the depression in the freezing point, relative to the freezingpoint of the 100% pure material, is directly proportional to the overallimpurity concentration divided by the "first cryoscopic constant." This is
expressed as:
T(pure) - T(obs) = x2/A. (36)
In eq (36), T(obs) is the observed freezing point of the particular sample beinginvestigated, T(pure) is the freezing point of the 100% pure material, x2 is themole fraction impurity concentration, and A is the first cryoscopic constant.A is given by the relation:
A = L/R[T(pure)] 2, (37)
28
Table 7. Latent heats of fusion and first cryoscopic constants of definingfixed-point materials
Substance Fixed PointTemperature
T/K
Latent Heatof Fusion
kJ/mole
•H2 13.8033 0.117
Ne 24.5561 0.335
o254.3584 0.444
Ar 83.8058 1.188
Hg 234.3156 2.292
H2 273.16 6.008
Ga 302.9146 5.585
In 429.7485 3.264
Sn 505.078 6.987
Zn 692.677 7.385
Al 933.473 10.79
Ag 1234.93 11.30
Au 1337.33 12.364
Cu 1357.77 13.14
First Cryoscopic Constant
K" 1
0.0739
0.0668
0.0181
0.0203
0.00502
0.00968
0.00732
0.00213
0.00329
0.00185
0.00149
0.000891
0.000831
0.000857
where L is the molar heat of fusion and R is the molar gas constant. (Note:
eq (36) is an approximation. A more complete expression includes secondarycryoscopic constants. The term "first cryoscopic constant" is used here fordistinction. Also, in some cases, the term "cryoscopic constant" refers to the
reciprocal of eq (37) and in other cases, to the effect of impurities per literor kilogram of solvent.) The first cryoscopic constants of metals arerelatively smaller than those of molecular substances and of the "cryogenic"gases (
3He, AHe , e-H2 , Ne , 2 , and Ar) . Referring to eq (36), zinc, which hasa first cryoscopic constant of 0.0018/K, requires an overall impurityconcentration of approximately 2 parts in 10 7 for the temperature of the half-frozen sample to be depressed by 0.0001 K, relative to the liquidus point. Onthe other hand, argon, with a first cryoscopic constant of 0.0203/K, requiresan impurity concentration close to 2 part in 10 6 for the same temperaturedepression. Table 7 lists the heats of fusion and the first cryoscopicconstants of substances specified for the defining fixed points. It is the
usual practice at NIST to calibrate SPRTs during the first 50% of the freeze.
3.2.2 TRIPLE POINTS OF e-H* . Ne .
g. AND Ar
3.2.2.1 GENERAL CONSIDERATION OF APPARATUS DESIGN
The cryogenic fixed points (triple-points of pure gases: equilibrium hydrogen,natural neon, oxygen, and argon) are best realized in a calorimetric typeapparatus designed for calibrating capsule- type thermometers (SPRTs, RIRTs
,
germanium resistance thermometers (GRTs), and others)[1,2,17,18,20,21,41,47,58,59,78,79,80] . During calibration of the thermometers,
29
TO SAMPLERESERVOIR
LIQUID NITROGEN
TO HIGH VACUUMAND ELECTRICALLEAD TERMINALS
PORT FOR FILLING
LIQUID NITROGEN
67 cm
LONG STEM SPRTWELLS (7)
HE GAS MANIFOLDTO SPRT WELLS
TEMPERATURECONTROLLED SHIELD
PASSIVE SHIELD
SAMPLE CELL SHIELD
SAMPLE CELL
BAFFLES
CAPSULE SPRTWELLS (6)
VACUUM CAN
SUPER INSULATEDDEWAR
Figure 4. A schematic drawing of the NIST argon triple-point apparatus forcalibrating seven long-stem SPRTs and six capsule SPRTs . Six long-stem SPRTwells surround a central SPRT well, which is large enough to accommodate a
holder for calibrating a capsule SPRT. At the bottom of the sample cell, sixcapsule SPRT wells are circularly arranged between the long- stem SPRT wells.
30
COPPERCAPILLARY
AUXILIARY ISOTHERMAL
/ SHELL, HELD WITH NUT
CELL HEATER
COPPER TEMPERINGSTRIPS
SAMPLE CELL
COPPER TUBES
HELICALLY GROOVEDCOPPER SLEEVE
PLATINUMRESISTANCETHERMOMETER
THERMOCOUPLECLIPS
Figure 5. A sealed cell suitable for containing cryogenic gases at highpressures. This cell of 20 cm3 volume was filled to 100 atmospheres with pureoxygen and used to realize the triple point of oxygen. Another cell of the sameouter dimensions, but of 16 cm volume and with wells for three capsule SPRTs
,
was filled to 163 atmospheres with oxygen and also used to realize the triplepoint of oxygen [47]
.
31
the calorimeter system thermally isolates the vessel containing the pure gas
sample and the thermometers. The capsule thermometers should be installed inclose-fitting wells of the sample vessel, using stopcock grease to enhance the
thermal contact. The wells should be vented for easy installation andsubsequent removal of the thermometers since such thermometers are sensitive to
shock. The wells can be vented by designing the well tubes to extend completelythrough the sample cell or by machining a small groove along the length of the
well wall. The number of wells in the vessel is limited only by the practicalsize of the vessel and of the calorimeter system. On the other hand, for largevessels, larger temperature gradients should be expected. The leads to the
thermometers should be tempered on the sample vessel. (Note: the term "temper"
refers to a process whereby sections of the leads or of the protective sheathsof long- stem SPRTs are placed in "steady- state thermal equilibrium" withselected parts of the apparatus.)
Since the triple point of argon is about 6 K above the normal boiling point(NBP) of nitrogen and since liquid nitrogen is readily available in largequantities, long-stem type SPRTs can be calibrated at the argon triple point in
an apparatus cooled by liquid nitrogen. The apparatus should be designed to
cool the upper part of the protective sheath of the long- stem SPRTs to liquidnitrogen temperatures, and then to heat the intermediate section of the sheathabove the sensing element to the argon triple point so that the sensing coilwill be in thermal equilibrium with the argon at its triple point [42] . See
figure 4 for a schematic of such an apparatus presently in use at the NIST.
Sample vessels for the cryogenic fixed points could be ruggedly constructed forhigh pressures and sealed with a suitable amount of the pure gas [41,79]. These"sealed-sample vessels" can be easily installed and removed from the calorimeterfor replacement of thermometers and also to be transported to other laboratoriesfor comparison (see fig. 5) . The amount of gas that can be sealed in suchvessels, however, is rather limited and, hence, the calorimetric system must beoperated with sufficient adiabatic control that the small amount of heat offusion of the sample is adequate to realize the triple point and then calibratethe thermometers. Also, since the amount of sample gas is small, extra caremust be taken to clean the vessel thoroughly before filling. The recommendedprocedure is to bake the vessel at high vacuum and then purge many times withthe sample gas before finally filling and sealing the cell [41,47,79].
The vessel for a cryogenic fixed point can be installed in the calorimeter andconnected by a small diameter tube to a source of pure gas [1,2,31,32,44,58,59].In this design, enough condensed liquid could be used to nearly fill the vesseland, thus, to provide an abundant supply of heat of fusion for calibration ofthermometers. The tube from the vessel must be connected to an "externalexpansion volume" of appropriate size so that when the system is at ambienttemperature the pressure is not excessive. The vessel could also be connectedto an external rugged container into which all or nearly all of the sample canbe transferred by cooling, and then contained by a high-pressure valve. Sinceunder these conditions the sample vessel would not be subjected to highpressures, it can be constructed of thin copper parts. With appropriate gas
handling and cleaning provisions, the same vessel could be used with all of the
reference gases, stored in separate external rugged containers, and the
thermometers calibrated at the fixed points of the gases without the necessity
32
of having to remove the thermometers from the vessel between fixed-pointdeterminations. On the other hand, the "sample vessel" can have separatechambers for each of the gases. In the latter design, separate tubes for the
gases must enter the calorimeter system. Although each of the chambers of a
"multi- chamber sample vessel" would be relatively small, the amount of sample
that could be condensed inside each chamber would still be more than that whichis normally used with the high pressure "sealed-sample vessel." Similar to the
procedure used in filling sealed- sample vessels, a thorough baking, pumping, andpurging procedure before filling should be followed with permanently installedvessels
.
3.2.2.2 REALIZATIONS OF THE TRIPLE POINTS AND THEIR APPLICATION TO CALIBRATION
The procedure for realizing a triple point of the cryogenic gases is to firstcompletely freeze the sample. If the triple point is being realized for the
first time with the apparatus, the sample should be cooled sufficiently belowthe triple point to determine the heat capacity of the system (sample, vessel,
and thermometers) from about 5 K to 20 K below (depending on the gas) to about20 K above the triple point, and to determine the heat of fusion of the sampleduring the same series of measurements [41,47]. (Note: check thermometers mustbe calibrated along with the test thermometers. The measurements on the checkthermometers will serve to guide the heating process during the calibration, as
well as to provide measurement statistics.)
After cooling to the required low temperature, the vessel should be thermallyisolated by placing it under continuous adiabatic control. Then the followingseries of measurements should be performed:
1. the equilibrium temperature should be observed with the checkthermometer,
2. a measured amount of electrical energy should be added to the
system,3. a new equilibrium temperature should be established and measured,4. steps 2 and 3 should be repeated until three heat capacity points
are obtained below the triple point,5. then, the sample should be completely melted by introducing a
measured amount of heat,6. next, the equilibrium temperature just above the triple point should
be measured,7. then, three additional heat capacity points should be obtained above
the triple point in accordance with steps 2 and 3.
From the knowledge of the eight equilibrium temperatures (four below the triplepoint and four above the triple point) and the measured amounts of electricalenergies added, the heat capacities of the system below and above the triplepoint, and the heat of fusion, are calculated. If Q joules are added to thesystem from an initial equilibrium temperature TL
just below the triple point to
heat the system to a final equilibrium temperature Tf just above the triplepoint, the heat of fusion L is:
L = Q - CS (T - TO - Ct (Tf
- T ), (38)
33
where Cs and C^ are the mean heat capacities in the temperature intervals in thesolid and liquid phases, respectively, and TQ is the triple -point temperature.
For calibration of SPRTs, the sample should be completely frozen for the secondtime and the temperature set at about 1 K below the triple point, the samplethermally isolated, and the equilibrium temperature measured. From theknowledge of the previously determined heat capacity of the system below thetriple point and of the heat of fusion, add enough electrical energy to meltabout 10% of the sample. If Ti is the initial equilibrium temperature justbelow the triple point, the required amount of electrical energy Q x to melt 10%of the sample is:
Q x= 0.1L + C8 (T - r^ (39)
Once the system comes to equilibrium, measure the resistances of all of thethermometers. Repeat the measurements at 20%, 40%, 60%, 70%, and 80% melted.If the sample is about 99.9999% pure, all measurements on each thermometerthroughout this melted range should agree to within 0.1 mK to 0.2 mK.
In using a temperature fixed point, one must make corrections for thehydrostatic head of the liquid and for the gas pressure on the definedequilibrium state. Table 6 gives the dT/dp for the defining fixed points of the
ITS-90, both in terms of the external gas pressure to which the fixed-pointmaterial is exposed and in terms of the column of liquid.
3.2.2.2.1 TRIPLE POINT OF EQUILIBRIUM HYDROGEN. 13.8033 K (-259.3467 °C)
Hydrogen gas samples of 99.9999% and higher purity are readily available. [The
first cryoscopic constant of hydrogen is relatively high (0.0740/K).Consequently, the liquidus point of an ideal hydrogen solution of 99.9999%purity would be approximately 0.01 mK lower than that of 100% pure hydrogen.Except for helium and deuterium, all other impurities would be either frozen or
in solution in very small amounts.] The commonly used catalyst for convertingortho hydrogen to para hydrogen is hydrated ferric oxide (Fe2 3 »H2 or Fe0»0H)
.
Other oxides of magnetic elements, either pure or mixed-metal, such as those ofchromium, nickel, cobalt, and neodymium, also have been used as catalysts forortho to para hydrogen conversion. The hydrated ferric oxide catalyst is
prepared by mixing at about 30 °C relatively dilute solutions (about 2 molal)of ferric chloride and sodium hydroxide, with only a slight excess of sodiumhydroxide, washing the resulting gelatinous Fe(0H) 3
precipitate thoroughly withdistilled water, air drying at 140 °C for 24 hours, vacuum baking at 110 °C for16 to 20 hours, and back-filling with hydrogen while the catalyst is still hot[7,104]. The catalyst is activated by flowing hydrogen through it for about 4
hours while the catalyst is maintained at a temperature of about 150 °C. Thesample vessel and ancillary components should be designed to permit the wholeof the hydrogen sample to come into contact with the catalyst at the equilibriumtemperature. See references [2,31,58].
3.2.2.2.2 TRIPLE POINT OF NATURAL NEON. 24.5561 K (-248.5939 °C)
Neon gas samples of 99.999% purity are commercially available. Samples ofhigher purity may be obtained by special arrangement with the supplier. The
34
Type A TypeB
Figure 6. Two types of triple point of water cells with wells for platinumresistance thermometers. The cells contain pure air-free waterthermometer wells are made of precision-bore tubing
The
35
principal impurities are CO, H2 , He, and N2 . The CO and the N2 impurities canbe frozen out by slowly flowing the sample through a coiled tube immersed inliquid neon; the H2 and the He impurities can be removed by freezing the neonsample in liquid hydrogen and pumping, with care so that the lighter isotopesof neon are not preferentially removed. [The first cryoscopic constant of neonis 0.0668/K. Consequently, the liquidus point of an ideal neon solution of99.999% purity would be approximately 0.1 mK lower than that of 100% pure neon.]The purified sample should be collected in a clean stainless -steel cylinder bycooling the cylinder in liquid hydrogen (or cooled with liquid helium) . Ifdesired, the purified sample can be collected directly in the cooled samplevessel. See references [1,79].
3.2.2.2.3 TRIPLE POINT OF OXYGEN. 54.3584 K (-218.7916 °C)
Oxygen gas samples of 99.999% purity are commercially available. Accuratechemical analysis of oxygen is difficult and, therefore, the claimed purity maynot be correct. Unknown or undetected impurities are chemically less reactivethan oxygen, e.g, the noble gases, and, in particular, argon, which forms aperitectic with oxygen. Samples of purity greater than 99.999% may be obtainedby special arrangement with the supplier. [The first cryoscopic constant ofoxygen is 0.0181/K. Consequently, the liquidus point of an ideal oxygensolution of 99.999% purity would be approximately . 6 mK lower than that of 100%pure oxygen. ] Careful preparation by thermal decomposition of potassiumpermanganate can yield samples of 99.9999% purity or better [47]. The oxygensample should be stored in clean stainless-steel cylinders. See references[32,47,59,79,80]
.
3.2.2.2.4 TRIPLE POINT OF ARGON. 83.8058 K (-189.3442 °C)
Argon gas samples of 99.9999% purity or better are readily available. The usualimpurities are the components of air, moisture, and hydrocarbons. [The firstcryoscopic constant of argon is 0.0203/K. Consequently, the liquidus point ofan ideal argon solution of 99.9999% purity would be approximately 0.05 mK lowerthan that of 100% pure argon.] To fill the sample vessel, the gas may be useddirectly or it may be dried first by slowly passing it through a coiled tubeimmersed in either liquid oxygen or a Dry Ice/ethyl alcohol mixture. Seereferences [ 17 , 41 , 42 , 59 , 79 , 80 ]
.
3.2.3 TRIPLE POINT OF WATER. 273.16 K (0.01 °C)
The triple-point temperature of water is assigned the value 273.16 K on theKelvin Thermodynamic Temperature Scale and also on the ITS -90. It is thereference temperature for resistance ratios in platinum resistance thermometry.The water used in preparing triple point of water cells is pure water ofnaturally-occurring isotopic composition. Figure 6 shows two commonly usedtypes of triple point of water cells.
Triple point of water cells are usually prepared from river water that has beenpurified by chemical treatment and distillation. River water is expected to
have concentrations of deuterium and the heavier isotopes of oxygen that arelower than that of ocean water. The extreme difference in the triple points ofnaturally- occurring water, including polar water, is given as 0.25 mK [100].
36
It is expected that differences among water triple-point cells of river waterwould be much smaller than 0.25 mK. (The isotopic composition differencebetween river water and ocean water [100] has been estimated to cause no morethan a 0.050 mK difference in the triple-point temperature.) While the basicmaterial is plentifully available, preparation of water triple-point cellsrequires a special effort [5,40]. Although the effect on the triple-pointtemperature is negligible, a trace of air always remains in most sealed triplepoint of water cells. When a cell at room temperature is gently inverted fromone end to the other and a sharp "click" is produced through the water hammeraction, the amount of gas in the cell will have negligible influence on the
triple -point temperature.
3.2.3.1 REALIZATION AND APPLICATION OF THE TRIPLE POINT OF WATER
In preparation for producing an ice mantle that is required for realizing thetriple -point temperature of water, the thermometer well of the cell is wipedthoroughly dry, sealed with a rubber stopper, and the cell placed in an ice bathto cool to a few degrees above the ice point. When the cell has been cooled inthis manner, an ice mantle of fairly uniform thickness can be obtained.Withdraw the cell from the ice bath, set it upright on a stand, and place onedrop of ethyl alcohol at the bottom of the well. Introduce small amounts ofcrushed Dry Ice into the bottom of the well and continue to do so until a thickmantle is formed at the bottom. Then, fill the well with crushed Dry Ice to the
water level of the cell. Continue to add crushed Dry Ice to the well so as to
maintain the level of Dry Ice at the water level. If the Dry Ice level becomeslow before more is added, the ice mantle may crack. If the cell were precooledas indicated above, a solid ice bridge may form at the water level. If such a
bridge forms, melt it immediately with heat from the hands while gently shakingthe cell. The solid ice bridge can completely seal the cell at the top and anysubsequent formation of ice could produce enough pressure to rupture the glasscell. When a mantle of approximately the desired thickness (4 to 8 mm) is
formed, stop adding Dry Ice, replace the cell in the ice bath with the wellopening slightly above the water surface of the ice bath, and leave the cellthere until all of the Dry Ice evaporates. Then, fill the well with ice waterand store the cell in an ice bath or ice pack for a day before using it. Whenthe ice mantle is frozen by using Dry Ice, a process that usually requires lessthan one hour, the strains in the ice cause the "triple -point temperature" to
be about 0.2 mK low. These are removed by letting the mantle anneal for oneday.
Other methods can be used also to prepare the ice mantle. With ethyl alcoholin the thermometer well, any "cold finger" technique can be used. Thistechnique includes successively inserting liquid nitrogen cooled rods, using a
closed-end tube containing crushed Dry Ice, or using a heat-pipe cooler. Thesemethods require more time to freeze the mantle, but the strain produced in the
ice will be less than those produced by the Dry Ice technique.
After the strains in the ice mantle have been relieved by storing the cell inan ice bath for at least one day, insert momentarily a glass rod into the wellin order to melt a thin layer of ice next to the well. This forms an ice -waterinterface immediately adjacent to the thermometer well. The test for this"inner melt" is made by giving the cell a rotatory impulse to determine whether
37
Figure 7
.
bath.An SPRT in a Type A triple point of water cell immersed in an ice
A -- platinum resistance thermometer, B -- heavy black felt to shield againstambient radiation, C -- polyethylene tube for guiding the SPRT into thethermometer well (the tube has a small hole near the top of the thermometer wellto allow water, but not ice, to enter the tube.). D -- water vapor, E --
borosilicate glass cell, F -- water from the ice bath, G -- thermometer well(precision bore glass), H -- ice mantle, I -- air-free water, J -- aluminumbushing with internal taper at the upper end to guide the SPRT into its close-fitting inner bore, K -- polyurethane sponge for cushioning the SPRT, L --
finely divided ice and water.
3&
the ice mantle rotates freely about the axis of the thermometer well. The outerice -water interface guards and thermally stabilizes the inner ice -waterinterface temperature that is measured with the SPRT.
Figure 7 shows a triple point of water cell immersed in an ice bath with an SPRTinserted into the thermometer well. An SPRT should be precooled in a glass tubeof water in an ice bath before it is inserted into the triple -point well so thatthe thickness of the water layer next to the thermometer well will not becomeexcessive. Also, the time required for the SPRT to come into thermalequilibrium will be shortened. Heavy felt cloth should be used to cover the icebath in order to prevent ambient radiation from entering the bath and reachingthe thermometer element, which otherwise would cause the thermometer to give aslightly high (erroneous) reading. A plastic foam cushion should be placed atthe bottom of the thermometer well to protect the well and the SPRT. Sincewater is a poor thermal conductor, a close fitting aluminum sleeve should beused to enhance the thermal conduction.
The thermometer current should be imposed immediately after insertion of theSPRT into the cell so that readings can be made under conditions of steady-state self heating. Five to ten minutes or longer may be required beforesteady- state conditions are reached. To avoid errors due to variations in theself heating that arise from variations in the thermal contact of thethermometer with its surroundings, it is best to read the SPRT at two currentsand extrapolate the readings to zero power in the SPRT.
3.2.4 FREEZING. MELTING. OR TRIPLE POINTS OF METALS: Hg. Ga. In. Sn. Zn. Al
.
Ag. Au. or Cu
The realization of metal fixed points requires the continuous presence ofliquid-solid or liquid-solid-vapor phases in thermal equilibrium. With SPRTs
,
the liquid-solid interface, i.e., the equilibrium whose temperature is measured,must surround and must be as close to the temperature sensing element as
possible. Since the first cryoscopic constants of metals are relatively low,
the fixed-point metal samples should be at least 99.9999% pure. Figure 8 showsidealized liquid/solid equilibrium conditions inside fixed point cells used in
freezing and melting experiments. Figure 9 shows a representative arrangementof an SPRT inserted inside a metal fixed point cell. Ideally, and similar to
the water triple-point cell, an outer liquid-solid interface, which completelysurrounds the inner interface, exists close to the container wall. This outerinterface, which has a temperature very close to that of the inner interface,thermally protects and thermally stabilizes the inner interface. In freezingexperiments, a layer of solid is first formed at the crucible wall, then a thinlayer of solid is induced on the thermometer well by inserting cooling rods.
As freezing advances, the outer interface approaches the inner interface untilall of the material is solid. In melting experiments, a layer of liquid is
first formed next to the crucible, then a thin layer of liquid is formed nextto the thermometer well by inserting a warming rod or a long heater. As meltingadvances, the outer liquid/solid interface approaches the inner interface.
Since different furnace or bath designs are required for fixed-point cellsoperated at different temperatures, they will be discussed along with each ofthe fixed points, or references will be made to appropriate sources of
39
/J
l-J~
*
'::
*
m t-'/
ri -K,-L
Cell Holder* »
Sample Cell
Sample
Liquid Sample
Solid Sample
ThermometerWell
ResistanceThermometer
ij; '\' SZi L
"38
-•<>'-;
'*B
Melting Freezing
Figure 8. Idealized liquid/solid equilibrium conditions inside fixed pointcells used in freezing and melting experiments. In freezing experiments, a
layer of solid is first formed at the crucible wall, then a thin layer of solidis induced on the thermometer well by inserting cooling rods. In meltingexperiments, a layer of liquid is first formed next to the crucible, then a thinlayer of liquid is formed next to the thermometer well by inserting a warmingrod or a long heater. As melting advances, the outer liquid/solid interfaceapproaches the inner interface.
40
Figure 9. An SPRT in a metal freezing-point cell.
A -- platinum resistance thermometer, B - - to helium gas supply and pressuregauge, C -- thermometer gas seal with silicone rubber, D -- silicone rubberstopper, E -- thermal insulation (washed Fiberfrax) , F -- thermometer guide tube[precision bore tube, ground (matt finish) to uniform outside diameter], G --
heat shunt (graphite) in close contact with F and with H, H -- borosilicateglass cell [precision bore tube ground (matt finish) to uniform outsidediameter], I -- graphite cap (lid) for the graphite crucible, J -- graphitethermometer well, K -- metal sample, L -- graphite crucible, M -- thermalinsulation (Fiberfrax paper) between the graphite crucible and the borosilicateglass cell.
41
descriptions.
In radiation thermometry, the liquid- solid phase of the metal fixed point mustcompletely surround the blackbody radiator cavity.
3.2.4.1 CONTAINER MATERIAL
Containers for the fixed-point metals must not contaminate the metal sample, andthe container material must be rugged enough to retain its integrity underthermal cycling between the temperature of use and ambient temperature. Thecontainer material preferably should be inert to air at temperatures of use; ifnot, e.g., graphite above about 200 °C, the container plus the fixed-pointmaterial must be protectively enclosed in an inert gas, such as nitrogen, argon,or helium, using either a borosilicate or fused silica glass envelope. Itshould be assumed that the fixed-point material itself will react chemicallywith air and, thus, the material must be protected.
3.2.4.2 METAL FIXED POINT DEVICES FOR CALIBRATING SPRTs
Figure 9 shows a representative fixed-point cell that can be used forcalibrating long-stem type SPRTs. In the following sections, individual typeswill be described.
3.2.4.2.1 MERCURY TRIPLE POINT
3.2.4.2.1.1 MERCURY SAMPLE
On the ITS -90, the triple point of mercury (equilibrium phase state of mercurysolid, liquid, and vapor phases) is assigned the value 234.3156 K (-38.8344 °C)
.
Depending upon the choice of container material and operating procedure, it maybe more practical to realize the mercury freezing point at one standardatmosphere, the value being 234.3210 K (-38.8290 °C) . Mercury samples withimpurity content of 1 part in 108 or less can be prepared by potassium hydroxideand nitric acid washings, followed by triple distillation [48]. The alkali andacid washings can be carried out by vigorously bubbling clean filtered airthrough the mixture of mercury and the alkali or the acid. To remove anyremaining oxidizable impurities, the first two distillations should be carriedout under reduced pressures with a fine stream of clean filtered air bubblinginto the mercury in the distillation container. The third distillation shouldbe done under high vacuum to remove the noble metals. With the high-puritymercury (99.999999%), both freezing and melting techniques give triple-pointtemperatures that agree to within i 0.1 mK over most of the liquid- solid range.
[The first cryoscopic constant of mercury is 0.00503/K. Consequently, the
liquidus point of an ideal mercury solution of 99.999999% purity would beapproximately 0.002 mK lower than that of 100% pure mercury.]
3.2.4.2.1.2 CONTAINERS FOR MERCURY
Any container material can be used with mercury that is sufficiently rigid anddoes not dissolve in, or chemically react with, mercury in the temperature range
of storage and application. The choice will depend upon whether the mercuryfixed-point cell is to be used at its triple point, at its freezing point, or
42
Hi
FT-
10 O.D.x 1 Wall
38 O.D.x 2 Wall
'-£.
LlU--TA\
Vf-r
TmCM
oT-CM
9.5 O.D.x 0.9 Wall
13
h*. in toen ^> ^
«» OJ
1 < 1
-X1
9.5 O.D.x 0.13 Wall
r- > *. r
9.5 O.D.x 0.9 Wall
38.1 O.D.x 1.7 Wall
1.0 Thick
-1.6 Thick
Borosilicate
GlassType 304
Stainless Steel
Dimensions in mm
Figure 10. Two mercury triple-point cells, one constructed of borosilicateglass and one of Type 304 stainless steel. The two small -diameter tubes at thetop facilitate the cleaning of the cells before filling and sealing. The glasscell is sealed by melting the small -diameter tubes, but the stainless steel cellis sealed by pinching flat the small -diameter tubes and electric -arc weldingthem, thereby severing them at the middle of the flat.
43
at either one.
3.2.4.2.1.3 METAL CONTAINERS
Mercury is capable of dissolving most metals, at least at the low levels ofimpurity content (<1 ppm) that are the permitted maxima for metal fixed points.Iron, nickel, chromium, and tantalum have been reported to be soluble at only1 ppm or less; hence, stainless steel is an adequate container for mercury inmost temperature standard applications [53,103]. (Note: the solubility ofnickel in mercury may be close to the limit of 1 ppm.) Mercury triple -pointcells can be prepared by pinching and welding the small diameter tubes used for
cleaning, evacuating, and filling (see fig. 10).
3.2.4.2.1.4 CONTAINERS OF GLASS
It is expected that some metal impurities in glasses [single metal oxide (e.g.,
fused silica glass) or mixed metal oxides (e.g., borosilicate glass)] or
"ceramics" can be leached out by mercury when the mercury is stored in them formany years. Traditionally, "soft glass" has been considered suitable forstoring mercury [48]; however, soft glass, without special treatment, may besusceptible to breakage when thermally shocked. Borosilicate glass and fusedsilica glass are more practical choices for mercury containers. Figure 11 showsa borosilicate glass mercury triple-point cell inside a stainless steel holder.
3.2.4.2.1.5 CONTAINERS OF PLASTIC
Organic plastics, such as polyethylene, polytetrafluoroethylene (Teflon), orpolytrifluorochloroethylene (Kel-F) , all of which are free of metals, can beused to contain mercury and be used at the mercury freezing point. A stainlesssteel holder similar to that used for the glass mercury cell and for the
stainless steel mercury cell (see fig. 11) or similar to the holder used for the
indium cell (see fig. 12) could be used as the external holder for a plasticmercury cell. Although plastic cells have not been used yet in preparingmercury fixed-point cells, it would be practical and desirable to use plasticcells for realizing the mercury fixed point. Since the vapor pressure ofmercury at room temperature is sufficiently high that mercury vapors can betransported under vacuum conditions , the vapors should be confined by anatmosphere of helium or other inert gases.
3.2.4.2.1.6 ASSEMBLY OF MERCURY CELLS
A purified mercury sample can be vacuum distilled into glass containers, withthe glass filling tube then sealed under vacuum with a flame [43] . The mercurysample may be vacuum distilled into stainless steel containers and the fillingtube pinched, and then cut and sealed using electric-arc welding techniques[43].
3.2.4.2.1.7 REALIZATION AND APPLICATION
When the total impurity content of a mercury sample is about one part in 10 8,
both freezing and melting techniques yield triple -point temperatures agreeingto within + 0.1 mK over most of the liquid- solid range. A dual -stage
44
VALVE ANDVACUUM
B
A = "O" ring tube seal
B = Thermometer well
C = Ethyl alcohol in well
D = Indium gasket seal
E = Insulation, rolled paper tissue
F = Stainless steel jacket
G = Tubular connection for
cleaning and filling
H = Insulation, tissue paper
rolled around (I) for centering
I = Copper foil cylinder
J = Borosilicate glass cell
K = Mercury
L = Thermometer cushion
(fused quartz wool)
M = Stand for mercury cell
N = Insulation (Aluminum
silicate wool)
Figure 11. A borosilicate glass, mercury triple-point cell in a stainless steelcontainer. A stainless-steel, mercury triple-point cell may also be mountedinside the stainless steel container. With the high-purity mercury sample, bothfreezing and melting techniques yield triple -point values agreeing to within± 0.1 mK over most of the liquid- solid range.
45
uCD
C•i-i
n)
vCoo
0)
Vw
to
[fl
CD
w
W4-1
•rH
X)C
4->
cfl
a
c
o
p-
MbOO4J
Ox:Pm
a)
n
bO•Hfa
46
refrigerator can yield a temperature near -40 °C and, hence, could be used forfreezing mercury, but a much simpler stainless steel vacuum enclosure placed ina Dry Ice/ethyl alcohol mixture (-78 C
C) can reduce the freezing rate of mercuryto give a freeze duration of about 10 hours or more with 2.2 kg of mercury, andthat is perfectly adequate. Figure 11 shows such a stainless steel enclosurethat has been used with a borosilicate-glass mercury cell at the NIST.
To start a freeze, fill the stainless steel enclosure that contains the mercurytriple -point cell with dry air and immerse it in a Dry Ice/ethyl alcohol bath.Fill the thermometer well with ethyl alcohol and insert therein an SPRT formonitoring the temperature of the cell. Usually, mercury supercools about 6 °C
in a borosilicate glass cell but only about 3 °C in a stainless steel cell.When recalescence is observed, evacuate the stainless steel holder. Remove themonitoring SPRT from the well and replace it with a thin-wall stainless steeltube that contains ethyl alcohol and that has been cooled in a tube of ethylalcohol immersed in a Dry Ice/ethyl alcohol bath. Insert successively into thestainless steel tube two or three liquid-nitrogen cooled glass rods, for about5 minutes each, in order to freeze a thin layer of mercury around the thermometerwell. The purpose of the stainless steel tube is to collect the frost that formson the rods when they are removed from the liquid nitrogen. Remove the stainlesssteel tube and replace it with the monitoring SPRT, which has been cooling inthe tube of cold ethyl alcohol. Switch on the thermometer measuring current.(Note: it may be necessary to refill the thermometer well with a small amountof cold ethyl alcohol from the Dry- ice cooled tube before the monitoring SPRTis inserted into the well. The well should be completely filled with ethylalcohol when the SPRT is in the well.) With the induced inner freeze around the
thermometer well, temperature equilibrium is reached in about 5 minutes. Afterthe resistance of the monitoring SPRT is read, other cooled SPRTs are
successively inserted into the mercury cell and calibrated. The final readingin a cell is made with the monitoring SPRT in order to check the extent of the
freeze. This final reading of the monitoring SPRT must agree with the initialreading to within ± 0.1 mK. See reference [43] for more details on the
calibration procedure at the mercury triple point.
3.2.4.2.2 MELTING POINT OF GALLIUM
The melting point of gallium is assigned a temperature of 302.9146 K (29.7646 °C)
on the ITS -90. Gallium of 99.99999% purity can be obtained commercially. Atsuch high purity, both freezing and melting techniques should yield liquid- solidequilibrium temperatures that agree to within ±0.1 mK. Since the metal expandsabout 3% on freezing, plastic containers, such as polyethylene, polypropylene,or polytetrafluoroethylene, are the most suitable. These are sufficientlyflexible at around 30 °C to accommodate the volume change in the gallium. In
assembling the gallium fixed-point cell, the supercooled metal can be poureddirectly into the container. A second more rugged container of Nylon, glass,
or stainless steel should enclose the flexible container so that the pressureof the inert gas over the metal can be controlled at one atmosphere or beevacuated to observe the triple po-int. A gallium fixed-point cell, consistingof an all plastic container, that is used at the NIST is shown in figure 13.
Since gallium supercools as much as 25 °C to 70 °C, depending upon the plasticmaterial that is in contact with it, the most convenient method of observing its
liquid-solid equilibrium temperature is the melting technique. [The first
47
-4.5 cm-|
Figure 13. An all-plastic gallium melting/triple -point cell. The triple pointis realized by using the melting technique. The cell is periodically evacuatedthrough the valve
.
A -- valve (Zytel) , B -- bath lid (Plexiglas) , C -
pumping tube (polyethylene), E -- cap (Nylon), F --
G -- case (Nylon), H -- thermometer well (Nylon), I
of the case (Nylon)
.
support rod (Nylon), D --
sample container (Teflon)
,
-- gallium metal, J -- base
48
cryoscopic constant of gallium is 0.00732/K. Consequently, the liquidus pointof an ideal gallium solution of 99.99999% purity would be approximately 0.01 mKlower than that of 100% pure gallium.
]
3.2.4.2.2.1 REALIZATION AND APPLICATION
In order to solidify the gallium metal in a fixed-point cell, initially in the
supercooled state (e.g., at room temperature), first insert successively two or
three liquid-nitrogen-cooled copper rods into the thermometer well of the cellto induce nucleation, and then place the cell in an ice bath for about one houror longer. The cell with the solidified gallium may then be placed in an oilbath at a temperature of about 40 °C to partially melt the sample to form anouter liquid- solid interface. To form a liquid- solid interface next to the
thermometer well (an inner melt) , the bath oil may be circulated through thethermometer well by pumping the oil through a tube placed in the well. Afterabout 20 minutes in the hot oil bath, about 25% of the gallium will be melted.The inner liquid- solid interface can also be prepared by using an electric heaterin the well. The amount of electric energy required, e.g., to form about a 1
mm shell of liquid around the thermometer well, can be calculated from the outerdimensions of the thermometer well and the heat of fusion of gallium (see table
7) . The gallium cell is then securely mounted, completely immersed so that the
thermometer well will be filled with the bath oil, in a stirred oil bath,controlled at a temperature about 10 mK above the liquid- solid equilibriumtemperature (melting point or the triple point). Also, the cell can immersedin a fairly close-fitting, oil-filled aluminum or copper block, controlled at
a temperature about 10 mK above the equilibrium temperature. The monitoringSPRT is heated and then inserted into the thermometer well of the gallium cell.Readings are taken after about 20 minutes of equilibration in the cell. Themonitoring SPRT is replaced in the cell with a preheated test SPRT andmeasurements on it are made after about 20 minutes. A number of SPRTs can besuccessively calibrated in the same "melt". When all of the test SPRTs havebeen calibrated, a final measurement in the cell is made on the monitoring SPRT.
This reading of the monitoring SPRT should agree with the initial reading to
within ± 0.1 mK. Also, measurements with different melts should agree to within± 0.1 mK. See references [16,26,65,68,94].
3.2.4.2.3 FREEZING POINT OF INDIUM
The freezing point of indium is assigned the value 429.7485 K (156.5985 °C) onthe ITS-90. Metal samples of 99.9999% purity and higher are commerciallyavailable. The freezing point of indium is at a sufficiently low temperatureto permit the use of containers of high temperature plastics[polytetrafluoroethylene (Teflon), polyimide/amide , and others], borosilicateglass, and stainless steel [4,69,74,92]. See figure 14 for an example of a
Teflon container used for indium at the NIST. As used with metals that freezeat higher temperatures, graphite can also be used with indium. The metal is
available in the form of small pellets, wire, and rods. Suitable amounts fora sample can be easily weighed into the container. [The first cryoscopicconstant of indium is 0.00212/K. Consequently, the liquidus point of an idealindium solution of 99.9999% purity would be approximately 0.5 mK lower than thatof 100% pure indium.
]
49
9 mm1 mm
2 mm
CAP (TEFLON)
0.005" SPIRAL GROOVE ALONGINSIDE OF THERMOMETER WELL(1 OR 2 THREADS/in.)
10 mm
5 mm CHA. HOLETHROUGH CAP
34.34 mm ID
44.34 mm 0018.70 mm 00
5 mm
CONTAINER (TEFLON)
THERMOMETER WELL(TEFLON)
5 mm
36.34 mm ID
40.34 mm OD
12.70 mm (0.500 in.) ID
14.70 mm OD
14 mm OD
mm
30 mm OD FUSEDSILICA TUBE
Figure 14. An all-plastic indium freezing-point cell to be used in a stainlesssteel container, such as that shown in figure 12. The argon gas pressure insidethe stainless steel container is adjusted to one atmosphere at the freezingpoint. A similar all-plastic cell and stainless steel container may be used to
realize the mercury freezing point or Belting point at one atmosphere.
50
3.2.4.2.3.1 REALIZATION AND APPLICATION
A tube furnace containing the indium-point cell is controlled about 5 °C abovethe freezing point of indium until the metal is completely melted. (It is
convenient to control the furnace temperature automatically and melt the metalsample overnight so that the freezing of the metal may be started in the
morning.) Insert the check SPRT into the cell well and when the SPRT indicatesthat the sample is about 5 °C above the freezing temperature, change the furnacetemperature control settings to control at 5
CC below the freezing point. When
the check SPRT indicates recalescence, change the furnace temperature controlsettings to control at 1 "C to 0,5 "C below the freezing point. Withdraw the
check SPRT from the cell well and insert successively in the well two fusedsilica glass rods, each initially at room temperature, for about 5 minutes eachand then insert the cool check SPRT in order to freeze a thin mantle around thethermometer well. (To avoid the consequence of inserting borosilicate glassrods into the aluminum or silver point cell to form the mantle around the
thermometer well, all glass rods used for this purpose in the laboratory shouldbe fused silica glass.) Within 20 to 30 minutes, the readings on the check SPRTshould indicate that the cell is at temperature equilibrium. After the readingson the check SPRT are completed, test SPRTs, that have been heated in anauxiliary furnace, are successively inserted into the cell well and calibrated.After all of the test SPRTs have been calibrated, the preheated check SPRT is
inserted again into the cell well and read. This second reading should agreewith the first to within ± 0.1 mK. See references [4,69,74,92].
3.2.4.2.4 FREEZING POINT OF TIN
The freezing point of tin is assigned the value 505.078 K (231.928 °C) . Metalsamples of 99.9999% purity are commercially available. Graphite containers arecommonly and successfully used for tin. Although the use of materials such as
boron nitride (BN) has not been reported, it could be a suitable container fortin. High purity tin has been found to supercool 25 °C or more [73,76]; hence,the freeze is nucleated by rapid cooling outside the furnace. The metal is
available in the form of small pellets and in rods suitable for filling thegraphite container. A method for filling graphite containers and installing the
graphite thermometer wells is described in reference [46] . [The first cryoscopicconstant of tin is 0.00329/K. Consequently, the liquidus point of an ideal tinsolution of 99.9999% purity would be approximately 0.3 mK lower than that of 100%pure tin.
]
3.2.4.2.4.1 REALIZATION AND APPLICATION
A tube furnace [46] containing the tin freezing-point cell is controlled about5 °C above the freezing-point temperature until the metal is completely melted.(It is convenient to control the furnace temperature automatically and melt themetal overnight so that freezing of the metal can be started early in themorning.) Insert the check SPRT into the cell well and, when the SPRT indicatesthat the sample temperature is about 5 °C above the freezing point, change thefurnace temperature control settings to control at 1
CC to 0.5 °C below the
freezing point. When the check SPRT indicates that the cell temperature is
close to the freezing-point value, withdraw the cell and the SPRT from thefurnace. The cell will then cool rapidly and when the SPRT detects recalescence,
51
replace the cell in the furnace. Withdraw the check SPRT from the cell well.
Insert successively in the well two fused silica glass rods, each initially atroom temperature, for about 5 minutes each, and then the cool check SPRT in orderto freeze a thin mantle around the thermometer well. Within about 20 to 30
minutes , the readings on the check SPRT should indicate that the cell is attemperature equilibrium. After the readings on the check SPRT are completed,test SPRTs, that have been heated in an auxiliary furnace, are successivelyinserted into the cell well and calibrated. After all of the test SPRTs havebeen calibrated, the check SPRT is heated and inserted again into the cell welland read. This reading should agree with the initial reading to within± 0.1 mK. See references [43,46,73,76].
3.2.4.2.5 FREEZING POINT OF ZINC
The freezing point of zinc is assigned the value 692.677 K (419.527 °C) . Metalsamples of 99.9999% purity are commercially available. High purity liquid zinchas been found to supercool about 0.02 °C to 0.06 °C; hence, unlike the freezingprocedure used with the tin-point cell, its freeze can be initiated in thefurnace without withdrawing the cell from the furnace. Graphite containers arecommonly and successfully used for zinc. Although the use of materials such as
boron nitride (BN) has not been reported, it could be a suitable container forzinc. The metal is available in the form of small pellets and in rods suitablefor filling graphite containers. [The first cryoscopic constant of zinc is
0.00185/K. Consequently, the liquidus point of an ideal zinc solution of99.9999% purity would be approximately 0.5 mK lower than that of 100% pure zinc.
]
3.2.4.2.5.1 REALIZATION AND APPLICATION
A tube furnace containing the zinc -point cell is controlled about 5 °C above thefreezing point until the metal is completely melted. If the furnace temperatureis maintained at a higher temperature, the zinc will melt faster, but the zincshould never be heated by more than about 5 °C above its melting point. (It is
convenient to control the furnace temperature automatically and melt the zincsample overnight so that the freezing of the metal can be started early in the
morning and the calibration of six or more test SPRTs can be completed duringthe same day.) Insert the zinc-point check SPRT into the cell well. When the
SPRT indicates that the melt is about 5 °C above the freezing point, change thefurnace temperature control settings to control at 5 °C below the freezing pointin order to initiate rapid cooling for nucleation. When the check SPRT indicatesrecalescence, change the furnace temperature control settings to control at 1 °C
to 0.5 °C below the freezing point. Withdraw the check SPRT from the cell welland insert successively into the well two fused silica glass rods, each initiallyat room temperature, for about 5 minutes each, and then insert again the coolcheck SPRT. This freezes a thin mantle around the thermometer well. Withinabout 20 to 30 minutes, the readings on the check SPRT should indicate that thecell is at temperature equilibrium. After the readings on the check SPRT arecompleted, test SPRTs, that have been heated in an auxiliary furnace, aresuccessively inserted into the cell well and calibrated. After all of the testSPRTs are calibrated, the preheated check SPRT is inserted again into the cellwell and measurements made on it. This second reading should agree with the
first to within ± 0.1 mK. See references [43,73,75].
52
51.3 cm
Figure 15. A graphite freezing point cell enclosed inside a fused silica tubewith tube connection to high vacuum, purified argon gas source, and pressuregauge
.
A -- connection to high vacuum, purified argon source, and pressure gauge, B -
- fused-silica-to-Kovar graded seal, C -- fused-silica connecting tube, outersurface matte finished to minimize radiation piping, D -- thermometer guidetube, E -- heat shunts (Inconel disks), F -- thermal insulation (Fiberfrax) , G -
- fused-silica outer envelope, H -- graphite lid, I -- graphite thermometerwell, J -- fused-silica thermometer well, K -- fused-silica fiber-woven tape forcushioning the graphite freezing-point cell inside the fused-silica enclosure,L -- metal sample, M -- graphite crucible, N -- fused-silica fiber pad forcushioning the thermometer, -- Fiberfrax paper liner.
53
3.2.4.2.6 FREEZING POINT OF ALUMINUM
The freezing point of aluminum is assigned, the value 933.473 K (660.323 °C) .
Metal samples of 99.9999% purity are commercially available. High puritygraphite containers have been used successfully with aluminum. High purityliquid aluminum has been found to supercool about 1 °C to 2 °C; hence, its freezecan be initiated in the furnace without withdrawing the cell from the furnace.Aluminum is highly reactive, particularly at elevated temperatures; liquidaluminum is capable of dissolving many metals. Liquid aluminum reacts withmoisture, forming the oxide and dissolving the hydrogen. The compounds Al AC 3 andaluminum oxycarbide have been found in aluminum samples cast in graphite at1000 °C. Because of the high chemical reactivity of aluminum, the graphite cellcontaining the metal must be completely protected by enclosing the cell in a
fused silica envelope (see fig. 15). The argon or helium gas that is used to
pressurize the freezing metal at one atmosphere must be thoroughly devoid ofmoisture, hydrogen, oxygen, hydrocarbons, and other substances that would reactwith liquid aluminum. The cell must not be heated more than 5 °C above the
aluminum freezing point. [The first cryoscopic constant of aluminum is
0.00149/K. Consequently, the liquidus point of an ideal aluminum solution of99.9999% purity would be approximately 0.7 mK lower than that of 100% purealuminum.
]
3.2.4.2.6.1 ASSEMBLY OF AN ALUMINUM- POINT CELL
High purity aluminum can be obtained in the form of shots or rods . Determinethe internal volume of the graphite container, taking into account thethermometer well. Determine the mass of liquid aluminum required to fill thecell to within 0.5 cm of the graphite lid. Weight out aluminum shots or cut andclean aluminum rods that correspond to this mass . The rods should be cut witha carbide tipped tool and cleaned by etching in a hot (about 200 °C) solutionconsisting (by volume) of reagent grade phosphoric acid (15 parts), sulfuricacid (5 parts) , and nitric acid (1 part) , and then carefully rinsing many timesin distilled water. Load the graphite crucible with the aluminum sample andthen slide it into an extra-long fused-silica test tube such as that shown infigure 16. Insert the test tube into the tube furnace and evacuate it. Whilecontinuing to evacuate the tube, set the furnace temperature to control at about5 °C above the melting point of aluminum. When the sample has completely melted,cool it to room temperature, while continuing to pump the tube. If aluminumshot, or rods of odd sizes, are used, the graphite cell will require severalloadings and meltings before the desired amount of total sample has been loadedinto the cell. When the graphite crucible is appropriately loaded with thesample, replace the silicone rubber stopper at the mouth of the extra- long testtube of figure 16 with the device for inserting the graphite thermometer welland lid (see reference [45] or fig. 17). Insert the test tube into the tubefurnace, evacuate it, and then fill it with high purity argon to a pressureslightly above ambient. Melt the aluminum sample and push the graphite well andlid into the cell. Cool the sample to room temperature, while maintaining theargon pressure in the test tube slightly above the ambient pressure. Finally,assemble the graphite cell containing the aluminum sample into the desiredfreezing-point cell configuration (see references [43,45,70] or fig. 15).
54
TO HIGH ~VACUUM
TO PURIFIED
ARGON SOURCE
Figure 16. A method for filling a graphite freezing-point cell by melting themetal sample in the graphite crucible. The required amount of sample is placedin the graphite crucible and the crucible is inserted into the fused silicatube. To protect the sample, the fused silica tube is evacuated or filled withan inert gas (e.g., purified argon) before melting the metal in a tube furnace.Depending upon the geometry of the sample, the melting of two or more batchesof sample may be required.
A -- Silicone rubber stopper, B
D -- metal sample.fused silica tube, C -- graphite crucible,
55
560 mm
153mm
219mm
560 mm
244mm
Figure 17. An apparatus for installing a graphite thermometer well and lid ina graphite crucible containing a molten metal sample.
A -- stainless steel pusher rod, B -- silicone rubber gas seal (permits linearmotion of the pusher rod A) , C - - inlet for purified argon gas that is used inpurging and maintaining positive pressure of the gas during the assembly process,D -- silicone rubber stopper, E -- stainless steel flange attached to the pusherrod for pressing against the graphite lid and thermometer well during assembly,F -- graphite lid for the crucible, G -- slit on the pusher rod (the two halvesspring outward to hold the graphite thermometer well and lid while melting the
metal sample), H -- graphite thermometer well, I -- fused-silica tube, J -- a
part of the fused-silica tube where its I.D. matches closely with the O.D. ofthe crucible and its lid so that the lid can be easily guided onto the openingof the crucible, K -- graphite crucible, L -- molten metal sample.
56
3.2.4.2.6.2 REALIZATION AND APPLICATION
SPRTs that are to be calibrated at the aluminum point and higher must have fusedsilica, sapphire, or ceramic insulation for the resistance element and itsextension leads. Such high temperature SPRTs should be handled by proceduresthat avoid thermally shocking them.
A tube furnace containing the aluminum-point cell is controlled about 5 °C abovethe freezing point of aluminum until the metal is completely melted. It is
convenient to control the furnace temperature automatically and melt the metalovernight so that the freezing of the metal can be started early in the morningand the calibration of two or three test SPRTs completed during the same day.
Insert the aluminum-point check SPRT stepwise into the cell well. Since theSPRT will cool considerably between the time it is withdrawn from any auxiliarypreheat furnace and inserted into the aluminum point cell, the SPRT is heatedin the section of the thermometer guide tube that is maintained close to thefurnace temperature. The SPRT is inserted initially to a location where its tip
is about 3 cm above the graphite cell lid. After about 5 minutes, the SPRT is
inserted an additional 5 cm and after another 5 minutes, another 5 cm, and so
on until the tip of the SPRT is at the bottom of the thermometer well. When the
SPRT indicates that the sample is about 5 °C above the freezing point, changethe furnace temperature control settings to control at 5 °C below the freezingpoint in order to initiate rapid cooling for nucleation. When the check SPRTindicates recalescence, change the furnace temperature control settings to
control at 1 "C to 0.5 "C below the freezing point. Withdraw the check SPRTstepwise from the cell, first to a location about 3 cm above the graphite celllid. After about 5 minutes, withdraw the SPRT another 5 cm. and after another5 minutes, another 5 cm, and so on until the SPRT is completely out of thethermometer guide tube. Next, insert successively into the cell well two fusedsilica glass rods, each initially at room temperature, for about 5 minutes eachto freeze a thin mantle of solid aluminum around the thermometer well. Againinsert the check SPRT stepwise, as described above, into the cell. Within 20
to 30 minutes, the readings on the check SPRT should indicate that the cell is
at temperature equilibrium. After the readings on the check SPRT are completed,the SPRT is removed from the cell stepwise, as described above. The test SPRTsare then successively inserted stepwise, calibrated, and removed stepwise as
described for the check SPRT. After all of the test SPRTs have been calibrated,the check SPRT is inserted again into the cell well and measurements made. Thissecond reading at temperature equilibrium should agree with the first to within± 0.1 mK.
When SPRTs are cooled rapidly from the aluminum point to ambient temperature,lattice vacancies are quenched in and these must be removed before the SPRTs are
calibrated at the triple point of water. To relieve any quenched in latticevacancies, the SPRTs that have been calibrated at the aluminum point should beheated in an auxiliary furnace at about 660 °C for about 30 minutes and thengradually cooled to about 500 °C over 3 hours or more before withdrawing fromthe furnace to cool at ambient temperature. See references [43,45,70].
3.2.4.2.7 FREEZING POINT OF SILVER
The freezing point of silver is assigned the value 1234.93 K (961.78 °C) . Metal
57
samples of 99.9999% purity are commercially available in the form of pellets.
High purity graphite containers have been used successfully with silver. Liquidsilver has been found to supercool not more than 0.5 °C; hence, its freeze canbe initiated in the furnace without withdrawing the cell from the furnace. Seereferences [3,15,72,98].
Oxygen is known to "dissolve" in liquid silver and lower the freezing point.Although the dissociation pressure of Ag2 is expected to be quite high at thefreezing point of silver, the lowering of the freezing point may be a combinationof the solution of Ag2 and of oxygen in liquid silver. (The dissociationpressure of Ag2 is given as 414 atmospheres at 507 °C.) In a graphiteenvironment at the freezing point of silver, a small amount of the oxygen willeventually react with the graphite; however, a newly prepared cell should bepumped at high vacuum at about 1000 °C for a week before back filling to oneatmosphere with purified argon, nitrogen, or helium. [The first cryoscopicconstant of silver is very small (0.000891/K) . Consequently, the liquidus pointof an ideal silver solution of 99.9999% purity would be approximately 1.1 mKlower than that of 100% pure silver.]
3.2.4.2.7.1 REALIZATION AND APPLICATION
The freezing-point cell of silver may be assembled by a procedure similar to
that used with aluminum. See section 3.2.4.2.6.1.
SPRTs that are to be calibrated at the silver freezing point must have fusedsilica, sapphire, or ceramic insulation for the resistance coil and its extensionleads. Such high temperature SPRTs should be handled by procedures that avoidthermally shocking them.
A tube furnace containing the silver freezing-point cell is controlled at about5 °C above the freezing point until the metal is completely melted. (It is
convenient to control the furnace temperature automatically and melt the metalovernight so that the freezing of the metal can be started early in the morningand the calibration of two or three test SPRTs completed during the same day.)Insert the silver-point check SPRT stepwise into the cell well. Since the SPRTwill cool considerably between the time it is withdrawn from any auxiliarypreheat furnace and inserted into the silver-point cell, the SPRT is heated inthe section of the thermometer guide tube that is maintained close to the furnacetemperature. The SPRT is inserted initially to a location where its tip is about3 cm above the graphite cell lid. After about 5 minutes, the SPRT is insertedan additional 5 cm, and after another 5 minutes, another 5 cm, and so on untilthe tip of the SPRT is at the bottom of the thermometer well. When the SPRTindicates that the temperature of the sample is about 5 °C above the freezingpoint, change the furnace temperature control settings to control at 5 °C belowthe freezing point in order to initiate rapid cooling for nucleation. When the
check SPRT indicates recalescence , change the furnace temperature controlsettings to control at 1 °C to 0.5 °C below the freezing point. Withdraw thecheck SPRT stepwise from the cell well, first to a location about 3 cm above the
graphite cell lid. After about 5 minutes, withdraw the SPRT an additional 5 cm,
and after another 5 minutes, another 5 cm, and so on until the SPRT is completelyout of the thermometer guide tube. Next, insert successively into the cell well
58
two fused- silica glass rods, each initially at room temperature, for about 5
minutes each in order to freeze a thin mantle of solid silver around the
thermometer well. Then, insert again the check SPRT stepwise, as describedabove , into the cell . Within about 20 to 30 minutes , the readings on the checkSPRT should indicate that the cell is at temperature equilibrium. After the
readings on the check SPRT are completed, it is removed from the cell stepwise,as described above. The test SPRTs are then successively inserted stepwise,calibrated, and removed stepwise as described for the check SPRT. After all ofthe test SPRTs have been calibrated, the check SPRT is inserted again into the
cell well and read. This second reading should agree with the first to within± 0.2 mK.
When SPRTs are cooled rapidly from the silver point to ambient temperature,lattice vacancies are quenched in and these must be removed before the SPRTs aremeasured at the triple point of water. To relieve any quenched in latticevacancies, the SPRTs that have been calibrated at the silver point are heatedin an auxiliary furnace at about 960 °C for 30 minutes or so and then graduallycooled to about 500 °C over 3.5 hours or more before withdrawing from the furnaceto cool at ambient temperature. In order to protect the platinum of the SPRTfrom contamination by diffusion of metals at high temperatures (above 660 °C)
,
the SPRTs should be enclosed in a platinum tube or other protective device.
3.2.5 CONTROL CHARTS OF CHECK THERMOMETERS
Control charts should be kept for all of the check thermometers associated withthe various fixed points. Each of these charts will usually consist of a
chronological graph of the check thermometer resistance value, R(X) , obtainedfrom measurements in the fixed-point cell X, and of the ratio,
W(X) = R(X)/R(TPW), of the resistance value R(X) to the resistance value, R(TPW),
obtained from measurements in a triple point of water cell. Presumably, suchcharts have been kept by those involved in precision thermometry. Entries onsuch control charts are made each time the particular fixed-point cell is used.
Since those involved in precision thermometry would have used triple point ofwater cells for their work based on the IPTS-68 and they would have used W(X)
,
the same control chart can be continued with the ITS -90. The reason, of course,is that the behavior of the fixed-point cells is independent of the scale. Notall of the fixed points of the ITS -90, however, are the same as those of theIPTS-68. The ITS -90 uses some of the IPTS-68 fixed points but it also uses otherfixed points. See reference [64].
Some metrologists may not have used W(X) as defined above, but may have used theW(t) of the IPTS-68. In that case, there will be a discontinuity in theircontrol charts involving W when they implement the ITS-90 and begin using W(X)
.
The magnitude of the discontinuity will simply be the ratioR(273.16 K)/R(273.15 K)
.
Control charts involving only R(X) as a function of time will not have anydiscontinuity due to the change in the scale. Depending on the temperature ofinterest, of course, there may be need to start additional control charts.
59
3.3 RADIATION THERMOMETRY
Above the freezing point of silver (1234.93 K) , the ITS-90 is defined in termsof Planck's radiation law. The values of temperatures r90 on the ITS-90 areobtained from the observed ratios of the spectral concentrations of the radianceLA of a blackbody at the wavelength (in vacuum) A at T90 and at the referencetemperature T90 (X) according to eq. (31).
Inside a closed cavity, the radiation densities at different wavelengths A dependonly upon the temperature of the cavity walls . When a practical radiator is
designed with a small hole in the wall to observe the radiation density at A,
there arises the question of how much the observed radiation departs from theblackbody radiation for a radiator design of a given geometry and material ofconstruction. There are numerous papers on the theoretical analysis of theemissivities associated with cavity geometry and construction materials anddescriptions of radiator designs that have been used in radiation thermometry[8,9,34,35,54,77 ,87,89,95] . The emissivities of cavities constructed of specularreflectors and diffuse reflectors have been analyzed [87] . It is expected thatat high temperatures many materials become oxidized and, consequently, becomediffuse reflectors. Although it is difficult to determine the actual temperaturegradients in a cavity, the effect of temperature gradients has also been treated[10,11]. The effective emissivity of a graphite blackbody cavity has beencomputed to be 0.99997 ± 0.00003 [77].
For radiation thermometer calibrations at the silver, gold, or copper fixedpoint, the blackbody cavity should be constructed of graphite and surrounded bythe freezing or melting metal contained in graphite to retain the high purityof the metal that is used. [The first cryoscopic constants of all three of thesemetals are extremely low (silver: 0.000891/K; gold: 0.000831/K; Cu: 0.000857/K).Consequently, the ideal-solution liquidus points of these metals of 99.9999%purity would be approximately 1.1 mK to 1.2 mK lower than that of 100% puremetals
.
]
References [27,55,56,57,62,77] give some details of construction of suitablegraphite fixed-point blackbody cavities. The metal should be protected from airusing an inert gas, such as argon, nitrogen, or helium, at a pressure slightlyabove ambient. The graphite container or auxiliary graphite scavengers canremove small amounts of oxygen impurities. For the blackbody cavity at theplatinum point, pure alumina has been used in an oxidizing atmosphere to avoidthe reaction between platinum metal and alumina in which oxygen gas is formedand metallic aluminum is dissolved in the platinum [33,86].
Usually, optical pyrometers or photoelectric pyrometers are used to determinethe ratio of the radiances of a source of unknown temperature with that of thereference source. The optical system of the instrument is designed to focus anearly monochromatic image of the radiation source onto a photodetector , whichuntil about the mid- 1950 's was only the human eye; now the eye has been replacedin high precision measurements by photoelectric detectors because of theirgreater accuracy and their suitability for automation of the measurements. Twomethods are commonly used to determine the ratios of spectral radiances. Eitherthe photoelectric pyrometer is designed for comparing the two radiation sourcesby null-detection operation, similar in principle to the disappearing filament
60
optical pyrometers using suitable neutral filters or sectored discs for
attenuating the radiation of one source, or for measuring directly the radiationdensity in terms of the detector output, e.g., photocell current. The latterrequires high stability and linearity of signal processing [34,77]. The opticsof the system may comprise refracting components (lenses) or reflectingcomponents (mirrors) [28,34,54,57,87].
Equation (31) requires the ratio of monochromatic radiances. Usually,interference filters are used for this purpose. The bandwidth should be narrowwith high transmittance while completely blocking out wavelengths outside the
desired band. The temperature error is smaller the narrower the bandwidth. Thetemperature of the filters should be controlled, since they are sensitive to
temperature changes. The photoelectric detector should be protected fromundesired radiations from outside the solid angle defined by the aperture of theblackbody cavity. Where the output of the photocell is used to determine the
ratio of the radiances, the linearity of the detector should be carefully checkedor calibrated.
For details of optical pyrometer operation and attendant sources of errors, see
references [ 28 , 34 , 54 , 57 , 77 , 87 ]
.
4. CALIBRATION OF THERMOMETERS ON THE ITS -90 AT VARIOUS LEVELS OF UNCERTAINTYAND SOME APPROXIMATIONS OF THE SCALE
In a standards laboratory, the design of apparatus and equipment for calibrationof thermometers on the ITS -90 should be based on the desired accuracy, the numberof thermometers and thermometric instruments that must be calibrated per year,the cost of realizing the ITS-90 (the fixed points and the measurementequipment), the cost of applying and maintaining the ITS-90, and the cost ofresearch to maintain and make necessary improvements on the realization of the
ITS-90. In a national standards laboratory, the efforts are directed toward theaccurate realization of the ITS-90.
On 1 January 1990, no laboratory was able to calibrate thermometers over the
complete range of the ITS-90 in accordance with the strict definition of the
scale. Also, it is thought that on that date there was no immediate, wide-spread requirement for "experimental calibration conversion" from the IPTS-68(75)to ITS-90 over the complete range. Since the differences between IPTS-68(75)and ITS-90 were known, "arithmetical conversions" should have met most of the
immediate requirements. Also, where stable thermometers have been used to
maintain the EPT-76 or parts of the IPTS-68(75), the scales on those referencethermometers could be converted to ITS-90, using the published approximatedifferences between the scales, and then those thermometers can be used to
calibrate other thermometers on the ITS-90. To realize the ITS-90 as definedand for international traceability , however, it is essential for the nationalcalibration laboratory to have all of the fixed-point apparatus and measurementequipment. Furthermore, without continued research and comparison with otherstandards laboratories, the question regarding the accuracy of the realizationof the scale will remain. The ITS-90 temperature calibrations are based on thethermal equilibrium states (vapor- liquid or liquid- solid equilibrium at knownpressures, or vapor-liquid-solid triple points) of pure substances. Substances,however, have some impurity content; the amount must be small enough to have
61
negligible effect on the measurement of temperature. Obviously, the fixed-
point device and the experimental procedure must be designed so that duringcalibration, the thermometer will be in thermal equilibrium with the equilibriumstate of the defining fixed point. A method for checking whether or not thethermometer is in thermal equilibrium with a metal fixed-point standard deviceis to reduce the immersion in the device a known amount or vary the experimentalconditions. The observed temperature change of the thermometer must correspondwith the hydrostatic head effect of the liquid metal in the device, or there mustbe no observed temperature change with experimental conditions (such as changingfurnace temperatures of metal fixed-point cells)
.
In order to determine the precision of the calibration process, it is essentialto use check thermometers with every calibration. The results of the checkthermometers will show whether the calibration process is "under statisticalcontrol" or not. The accumulated results show the precision of the "gross"calibration process.
Since some parts of this section deal with approximations of the ITS -90, andwill make reference to the scale differences given in table 1, the methods bywhich the table was constructed will be described. The differences (T90 - T76 )
between 5 K and 27 K were obtained using the same relation [99] as that used for
(rNPL-75 " T76). namely,
(T9Q - r76)/mK = - 0.0056(r90/K)2
. (40)
The differences (r90 - T68 ) between 14 K and 100 K were obtained by Working Group4 of the CCT by graphical interpolation of data from the published literature.The differences (T90 - T68 ) , or (t90 - t68 ) , between -200 °C and 630 °C wereobtained by Working Group 4 of the CCT from published data on two SPRTs, oneSPRT covering the range below °C, and the other covering the range from °C
to 630 °C. A polynomial of the form:
8
(t90 - t68)/°C = I a^t^A^O) 1. (41)
i=l
was fitted to the data from -200 °C to 630 °C and the coefficients are:
a.x= -0.148759 a5
= -4.089591a2 = -0.267408 a6
= -1.871251a3= 1.080760 a7
= 7.438081aA
= 1.269056 a8= -3.536296
The polynomial with these coefficients reproduce the tabulated differences [83]to within 1 mK above C
C and to within 1.5 mK below °C. The differences(t90 - t68 ) between 630 °C and 1064 °C were obtained by/ Working Group 4 bygraphical interpolation from published data [14]. The differences (t90 - t68 )
above 1064 °C were obtained from the equation:
(tgo - t68 )/°C = - 0.25 [<t90/°C + 273.15)/(1337.33)] 2. (42)
In section 3, we discussed the direct realization of the ITS -90, using the
62
standard instruments of the scale, i.e., the realization of the scale at the
lowest level of uncertainty. Of course, even the standard interpolatinginstruments used at the same thermodynamic temperature will indicate temperaturesthat differ slightly due to the devices having nonideal behavior and the scalebeing expressed in as simple a form as possible. The differences in indicatedtemperatures, however, are negligible for all practical purposes, being of the
order of < ± 0.5 mK for temperatures above about 5 K (i . e ., assuming no errorsin calibration). The realization of the ITS-90 in the liquid helium range oftemperatures (0.65 K to 5.0 K) through helium vapor pressure -temperaturerelations can be accurate to about ± 0.1 mK or ± 0.2 mK.
For nonstandard types of thermometers used to approximate the ITS-90, the levelof uncertainty is higher than the numbers just given because of the inherentinstability of these thermometers. In all cases, these types of thermometersare calibrated by comparison with one or more standard instruments of the scale,
e.g. , vapor -pressure thermometry; vapor -pressure thermometry and gas thermometry;vapor-pressure thermometry, gas thermometry, and platinum resistance thermometry;gas thermometry and platinum resistance thermometry; platinum resistancethermometry; or pyrometers or spectral radiometers.
The NIST offers calibration services for various thermometers and pyrometerscovering the range from 0.65 K to 4200 °C (see NIST SP 250). Of this range, theChemical Process Metrology Division offers calibrations for contact thermometerscovering the range from 0.65 K to 2400 K, and the Radiometric Physics Divisionoffers calibrations for non- contact thermometers (radiation pyrometers) coveringthe range from 1234.93 K (961.78 °C) to 4200 °C. Calibrations of only contact-type thermometers will be discussed here. The types of contact thermometerscalibrated include rhodium-iron resistance thermometers (RIRTs)
,germanium
resistance thermometers (GRTs) , standard platinum resistance thermometers(SPRTs) , thermocouples (t/c) , liquid- in-glass thermometers, thermistorthermometers, industrial platinum resistance thermometers (IPRTs) , digitalthermometers, and other special thermometers that are compatible with the NISTcalibration equipment.
4.1 RHODIUM- IRON RESISTANCE THERMOMETERS
At temperatures below 13.8033 K, RIRTs and GRTs are, at the present time, the
only thermometers that are suitable for precision temperature measurements.Also, RIRTs (and to a lesser extent GRTs) are suitable for use at temperaturesup to the triple point of neon (24.5561 K) . In the range from 0.65 K to about25 K, RIRTs have reproducibilities of about ±0.2 mK. Consequently, RIRTs don'tdegrade the realization of the ITS-90 significantly.
When the ITS-90 is realized, as defined, at NIST, some NIST RIRTs will becalibrated at many temperatures through the use of vapor-pressure thermometryand gas thermometry to produce reference -standard RIRTs, which will beperiodically recalibrated. The resistance -temperature data of these RIRTs willbe represented by a polynomial.
Customer RIRTs are calibrated by comparison with reference -standard RIRTs. Apolynomial is fitted by the method of least squares to the RIRT resistance-temperature data so obtained and the results are reported in terms of the
63
polynomial that is selected.
Until the NIST completes the development of the CVGT and vapor-pressurethermometry apparatus with which the reference -standard RIRTs will be calibrated,calibrations of customer RIRTs are performed by comparison against reference-standard RIRTs that have been calibrated on the NPL-75 Scale [13] or on the
EPT-76 and converted to the ITS-90.
To convert a calibration of an RIRT on the EPT-76 to an approximate calibrationon the ITS-90, use the EPT-76 calibration resistance -temperature data, changethe T76 values to T90 values using the (r90 - T76 ) differences given in table 1,
or calculated with eq (40), to produce a new set of resistance-temperaturevalues, and then fit a polynomial of the required degree to these data. Usingthe coefficients of the polynomial so determined, produce the desired calibrationtable. A typical calibration report on the EPT-76 and one on the ITS-90 aregiven in appendix 3, sections 6.3.6 and 6.3.7, respectively.
4.2 GERMANIUM RESISTANCE THERMOMETERS
GRTs are comparable with, but not quite as stable as, RIRTs. They are calibratedin a manner similar to that of the RIRTs and their results similarly reported.At NIST, customer GRTs are calibrated by comparison with reference -standardRIRTs
.
Anyone with a calibration on the EPT-76 may convert it to an approximate ITS-90calibration by the same procedure as just outlined for RIRTs.
4.3 STANDARD PLATINUM RESISTANCE THERMOMETERS
Both capsule and long- stem type SPRTs are calibrated at the NIST. They will bediscussed separately.
4.3.1 CAPSULE SPRTs (13.8033 K TO 429.7485 K OR 505.078 IP
For temperatures in the range from 13.8033 K to 273.16 K, capsule SPRTs are themost suitable thermometers. NIST has reference capsule SPRTs that have been,or will have been, calibrated at the defining fixed points in this range. ThoseSPRTs are used in calibrating customer thermometers by the comparison techniqueover the range from about 13 K to 84 K. The temperatures at which comparisonsare made are at, or within a few mK of, the defining fixed-point temperaturesof the ITS-90 and at temperatures approximately mid-way between the fixed-pointvalues. Data at the temperatures intermediate to the fixed-point values areincorporated as a check on the calibration process. At and above the argontriple point (83.8058 K) , customer capsule SPRTs are calibrated by the fixed-point method. See section 6.3 for an example of how to calculate the
coefficients of the deviation functions.
4.3.2 LONG-STEM TYPE SPRTs (83.8058 K TO 1234.93 K)
Long- stem type SPRTs are used in the range from 83.8058 K to 1234.93 K. Twodifferent long- stem type SPRTs are required to cover this whole range, one type
being the customary SPRT having a nominal 25.5 resistance at °C and used in
64
the range from 83.8058 K to 692.677 K or to 933.473 K, and the second type havinga somewhat longer stem and having a nominal 0.25 Q or 2.5 resistance and usedin the range from 273.15 K to 1234.93 K. Such thermometers are very stable ifhandled carefully. In this range of temperature, SPRTs are calibrated by the
fixed-point method, different sets of fixed points being required for differenttemperature subranges. The required fixed points for the different subrangeswere specified in section 2 of this document. In the fixed-point method,corrections for hydrostatic heads and gas pressures over the metals of the
freezing-point and melting-point cells are made (see table 6). Similarly,corrections for hydrostatic heads present in metal and in gas triple -point cellsare made. As a check on the accuracy of the calibration, measurements are madeat one or more "redundant" defining fixed points lying within the temperaturerange of calibration or at a well characterized secondary fixed point, such as
the cadmium freezing point, that lies within the temperature range ofcalibration. The value of the temperature of a check point calculated from thecalibration constants should agree closely with the accepted value of that point.If not, then either an error was made in the calibration, one or more fixed-point cells are defective, or the thermometer is defective.
The procedures indicated in section 3.2 for realizing the various fixed-pointtemperatures and for handling SPRTs are followed carefully during calibrationof platinum thermometers, especially so for calibration at the aluminum andsilver freezing points (see sec. 3.2.4.2.7.1 for a discussion of latticedefects). See section 6.3 for an example of how the coefficients of thedeviation functions are calculated from the data for the two sets of fixedpoints
.
4.3.3 CONVERSION OF THE IPTS-68 CONSTANTS AND W(Tei ) TABLES TO APPROXIMATEITS-90 CONSTANTS AND W(Toq ) TABLES
For SPRTs that have been calibrated recently on the IPTS-68(75) , theircalibrations may be converted to approximate calibrations on the ITS-90 attemperatures between the triple point of equilibrium hydrogen (13.8033 K) andthe freezing point of zinc (692.677 K)
.
To make the conversion, first obtain values of W(T68 ) , i.e., R(T6a )/R(0 °C) , fromthe IPTS-68 (75) calibrations at values of T68 corresponding to the temperaturesof the relevant fixed points of the ITS-90 in the appropriate range in which theconversion is desired. [Note: for a fixed point, the temperature values r68 andT90 are different, but the "hotness" is the same and so the resistance of a givenSPRT remains unchanged. The temperature values on the two scales are definedto be the same only at the triple point of water and at the absolute zero of thetemperature scales. Due to the nature of the scales, however, there are othertemperature values of these scales that are also the same, which are fortuitous.See fig. 1, which shows the difference of t90 and t68 as a function of t90 .]
Using these values of W(T6B ) at the appropriate ITS-90 fixed-point temperatures,calculate values of W(TQQ ) , i.e., R(T90 )/R(273 . 16 K) , by dividing the values ofW(T68 ) at the appropriate fixed-point temperatures by the value of P/(273.16 K)
,
i.e., the value of W(T68 ) at the triple point of water. Table 8 shows samplesof such conversions for a capsule- type SPRT and for a long- stem type SPRTcalibrated on the IPTS-68(75). The calibration constants of the IPTS-68(75)equations for the two SPRTs are given also in table 8. The values of W(TQQ ) so
65
Table 8. Example of conversion of calibrations of SPRTs on the IPTS-68 to
approximate calibrations on the ITS -90
Capsule SPRT. S/N 1812284 : e-H2 TP to Sn FP
IPTS-68 CALIBRATION CONSTANTS
R(0 °C) = 25.4964808Aj. = -1.6314209xl0" 4
D x= 1.4525327xl0" 8
A2= -2.1616630xl0" 4
D2= -2.2572375xl0" 10
A3= -4.5449999xl0" 4
A4= 3.0860000xl0" 7
a = 3.9262754xl0" 3
B x= -8.7377446xl0" 7
B2= -1.7113512xl0" 6
7.7242433xl0"6
S = 1.4964667Cj = -4.0321350xl0"7
C2 = 6.6163769xl0"8
C 3= -3.9028516xl0"8
C A = -1.0880791x10"*
ITS -90 CALIBRATION CONSTANTS CONVERTED FROM THE IPTS-68 CONSTANTS
R(273.16) = 25.4974973a x
= -2.5239001x10"*a9
= -2.5287142x10"*b 1
= -1.2277862x10"*b Q = -1.1130131xl0" 5
Ci =
C2 =c 3 =c A
=
C, =
= -2.3783015xl0-6
= -4.3892024xl0" 6
= -1.5608728xl0"6
= -2.1374663xl0"7
= -1.0344171xl0"8
Long- Stem SPRT. S/N RS8YA-5 : Ar TP to Zn FP
IPTS-68 CALIBRATION CONSTANTS
R(0 °C) = 25.5086208
A4 = 2.6581418xl0_1 *
A = 3.9856609xl0" 3
a = 3.9268986xl0- 3
= -5.8762238xl0" 7
= 1.49640322= 9.3183900xl0" 7
ITS -90 CALIBRATION CONSTANTS CONVERTED FROM THE IPTS-68 CONSTANTS
R(273.16 K) = 25.5096386a4
= -9.3225823xl0" 5
a8= -9.1058813xl0" 5
b 4= -9.9914440xl0'6
bo = -7.6061559xl0" 6
obtained may then be used with the appropriate relations described in section2 to obtain the constants of the ITS -90 deviation equations for the SPRTs. Intable 8, we list the values of the deviation constants that were calculated fromthe sample data presented there. For comparison, values of the conversion from&f(r68 ) to W(T90 ) , as well as measured values of W(TQ0 ) , for the various relevantfixed-point temperatures T68 and Tgo , respectively, are given in table 9. Thedeviations in the values of the last two columns of table 9 for the capsule SPRTreflect the inconsistency between the NBS-IPTS-68(75) wire scale and thedifference between the IPTS-68 (75) and the ITS -90 as given in table 1. Notethat for the long- stem SPRT, the zero deviations for tin and zinc are a
consequence of those fixed points having been used in both cases and the
66
Table 9. Values of W(T68 ) and W(Tgo ) for various fixed-point temperatures T6B
rgo , respectivelyand
Capsule SPRT (S/N 1812284)
Fixed Point T68/K W(T6q) meas
.
r90/K W(T90 ) calc* W(T90 ) meas.
e-H2 TP 13 ,81 0.00119822** 13.8033 0.00119817 0.00119721e-H2 BP 17 .04200 0.00231120** 17.0357 0.00231111 0.00231049e-H2 BP 20 .280 0.00425962** 20.2711 0.00425945 0.00425815
Ne TP 24 .5616 0.00848337** 24.5561 0.00848303 0.00848391
2 TP 54 .361 0.09182547** 54.3584 0.09182180 0.09182102Ar TP 83 .79723 0.21598575** 83.8058 0.21597714 0.21597795Hg TP 234 .3086 0.84421212** 234.3156 0.84417846 0.84417781H2OTP 273 .16 1.00003987 273.16 1.00000000 1.00000000Ga MP 302 .9218 1.11815352** 302.9146 1.11810895 1.11810896In FP 429 .7848 1.60970773** 429.7485 1.60964355 1.60964566Sn FP 505,.1181 1.89263856
Lone- Stem SPRT
505.078
CS/N RS8YA-5^
1.89256311
i
:
1.89256311
Ar TP 83.,79723 0.21592943** 83.8058 0.21592084 0.21592281Hg TP 234.,3086 0.84418993** 234.3156 0.84415637 0.84415625H20TP 273.,16 1.00003976 273.16 1.00000000 1.00000000Ga MP 302. 9218 1.11817175*** 302.9146 1.11812729 1.11812699In FP 429.,7848 1.60980450*** 429.7485 1.60974050 1.60974062Sn FP 505.,1181 1.89278557 505.078 1.89271033 1.89271033Zn FP 692.,73 2.56885786 692.677 2.56875573 2.56875573
Values of W(TQ0 ) calc. were obtained by conversion of corresponding values ofW(T68 ) meas.
These values were calculated from NBS-IPTS-68 calibration constants, based onthe NBS wire scale and calibration at the triple point of water and the freezingpoint of tin (and the freezing point of zinc for the long-stem SPRT)
.
**These values were calculated from IPTS-68 calibration constants determinedfrom fixed points.
conversion process having been consistent. The deviations for the other fixedpoints (Ar, Hg, Ga, and In) reflect the non- uniqueness of this SPRT and possibleerrors incurred in measurements at those fixed points.
The results of tables 8 and 9 for the long- stem SPRT, with regard to thedifferences between temperatures determined from an actual ITS -90 calibrationand those determined from an IPTS-68 (75) calibration but then calculated forapproximate ITS -90 values, are shown in figure 18. Note that at temperaturesabove 273.16 K, the zero deviation of figure 18 is a consequence of the fact thatthe same fixed points (tin and zinc) have been used in both cases and theconversion process having been consistent.
67
c
•H•p(d
uA•H
(0 1u o
CO in<o •
I inco CM
B „
H in
-0 rtj
C >"(0 CO
en
C P
•H4J C(0 •HM J3ja UHH »(0 ou G\
o U)0\ Eh1 HCOH oH +J
C -00) 0)
0) +J
s np 0)
0) >CQ G
OJ uUCd)
uCO
M-l
<W•HQ
in
o o*
oCM
oHO
ooI
oinvo
oovo
oinin
ooin
om
oo
oinn
oo
oin
ooCM
om
oo
oin
«
a)
u
+J(0
u
a)
Eh
in
o o o ooo o
I
CO\o
oenEhI
oCKEH
ooCO
525
CO
Hodp-i
C/3
sCD
4J
w
co
co
•p
oCM
C
a>
Eh
ca)
a>
>4->
a)
xi
CQ
CO
oc<u
co4-1
iw •
•m oo
°CO
• rH00 ,Qi-H cd
4->
CO
U <4-|
3 ObO•H -
68
4.3.4 UNCERTAINTIES OF CALIBRATIONS AND THEIR PROPAGATION
Both systematic and random errors of measurements introduced in a calibrationare propagated throughout the temperature range of the calibration. It is the
total uncertainty arising from both of these types of errors, however, that is
of interest to the customer and user of a calibrated thermometer. Internationalcomparisons of fixed-point cells below 90 K [79] and other data suggest that the
uncertainties (at the la level) in the realizations of the defining fixed pointsof the ITS -90 are about ± 0.2 mK for the triple points of hydrogen and neon,± 0.1 mK for the triple point of oxygen through the melting point of gallium,± 0.7 mK at the freezing point of indium, ± 1 mK from the freezing point of tinthrough the freezing point of aluminum, and i 2 mK for the freezing point cfsilver.
The uncertainty of temperature measurements in the liquid helium range (0.65 Kto 5.0 K) results from the uncertainty of the helium vapor-pressure measurements.The uncertainty (at the la level) throughout this range of temperature is
estimated to be approximately ±0.1 mK to ± 0.2 mK.
For the CVGT, the uncertainty in the measured temperature over its temperaturerange (3.0 K to 24.5561 K) arises from uncertainties of realizations of Che
triple points of neon and of equilibrium hydrogen, of the measurement of theCVGT gas pressure, and of the measurement of the vapor pressure of helium. Theuncertainty (at the la level) throughout this range of temperature is estimatedto be approximately ± 0.1 mK to ± 0.2 mK. Uncertainties introduced by a
particular CVGT design may add to these uncertainties.
In the calibration of SPRTs over any given subrange, the possible error in therealization of each of the fixed points and any error of measurement must beconsidered and they will be propagated independently of that incurred at theother fixed points involved. The total uncertainty of measurements at a giventemperature is then the root-mean- square of the appropriate contributinguncertainties. Curves showing the propagation of a i (unit error) incurred at
each of the defining fixed points of the two major ranges are shown in figures19 and 20. Figure 19 shows the propagation of errors associated with the fixedpoints below 273.16 K; and figure 20 shows curves for the fixed points used in
the calibration of SPRTs in the range from 273.15 K to 1234.93 K. The labelson the various curves indicate the fixed-point in which there is the unit errorin its temperature, and the other fixed points without error. As an example,consider the curve labelled Sn(Ag, Al, Zn) of figure 20. The symbol Sn indicatesthat the unit error occurs in the Sn freezing-point temperature; that is clearlyindicated by a unit offset of the curve at that point. The symbols in
parenthesis, i.e., (Ag, Al , Zn) , indicate that the Ag, Al , and Zn freezing pointswere the other fixed points involved in the calibration and that measurementswere made at those fixed-point temperatures without error. Also, it is assumedthat the measurements at the triple point of water were without error. The otherlabelled curves are similarly interpreted. The straight lines labelled TPW showthe errors propagated for an error of ± 0.1 mK made by the user in the triplepoint of water.
69
4.3.5 ESTIMATES OF POSSIBLE ERRORS INTRODUCED BY EXTRAPOLATIONS BEYOND THERANGE OF CALIBRATION
It is unwise and is poor practice to use any thermometer beyond the temperaturerange over which it was calibrated. Nevertheless, some users persist in doingjust that. In some cases, especially if the extrapolation is for only a fewkelvins, the error introduced may be rather small. The errors of some typicalextrapolations, calculated and depicted for NIST SPRTs , are shown in figures 21,
22, 23, 24, and 25.
The curve in figure 21 shows the error introduced for a NIST SPRT byextrapolating the deviation function, determined from calibration over the rangefrom the triple point of argon to the triple point of water, downward from theargon triple point to 54 K. Extrapolating downward to about the boiling pointof nitrogen (77 K) results in a fairly insignificant error in this case and,
thus, can be done with the usual caution that some thermometers may not be as
good as that of figure 21.
Figure 22 depicts the results for another NIST SPRT. The curve shows the errorintroduced by extrapolating the deviation function, determined from calibrationover the range from the triple point of mercury to the melting point of gallium,downward from the mercury triple point to 84 K.
Figure 23 displays the results for the same NIST SPRT as used for figure 22, butthe curve in this case shows the error introduced by extrapolating the deviationfunction, determined from calibration over the range from the triple point ofmercury to the melting point of gallium, downward from the mercury triple pointto only 200 K. A sizable error is shown for observations that would be made atDry Ice temperatures (-78 °C)
.
Figure 24 depicts the results for several NIST SPRTs, but the curves in thiscase show the errors introduced by extrapolating their deviation functions,determined from calibration over the range from the triple point of water to thefreezing point of zinc, downward from the triple point of water to -50 °C.
Figure 25 displays the results for several NIST SPRTs, but in this case thecurves show the errors introduced by extrapolating their deviation functions,determined from calibration over the range from the triple point of water to thefreezing point of zinc, upward from the freezing point of zinc to 934 K (660 °C) .
As a general rule and good practice, one should never extrapolate any of the
ITS -90 deviation functions beyond their range of application. If, however, theestimated uncertainty introduced by extrapolating a deviation function beyondthe range of calibration is acceptable, then the user may do so, but with the
knowledge that his particular thermometer may yield results with a largeruncertainty than that estimated. The user that makes such extrapolations shouldrealize that not all SPRTs will behave as indicated in figures 21 through 25.
The results depicted in these figures are examples only and are valid for onlythose SPRTs indicated. Other SPRTs may be better or worse.
70
01
uouu &W -U
a)
c eoH
4J eId Qj
M X!ja H•HH a)
id
u cIT!
<H PW•H
c CO
o 0)
•H «•p(0 gOi rj
ro ca -HP
u id
a, rHCM
oa\ ci •HtoEn
oCM
in oin o • •
o Ho o i 1
CD 4-1 - C T)43 o C CU 13
H a) 0) cd
u bO^Dn O en
J,<Sh <u TH
S^S-l
xs > CO
1) >s CfJ a;
r-l43 43 43
4J
ro(1)
> s o4-> a)
r^ •r-l•H S-i
CM 4-1
cd
bOa)
c
S-l X) cd
in
cm
X>ccd
43•Hr-l
•i-l
CD
(0
aa)
43
W 4J3 CO
rr)
U
oinCM
COCOO
•H
3 4-1 •I-l
CD
•H
00 CJ CO
Sh S-l Uro o CO as 3r-
1
4J +j oin
1) C cfl
CM(N
a)>r-l
•Ho
s S-i
o0) 4-> P-4-1 S-i
!S •r-l5-1
4J wiCD <D
o as n i—
1
4->
oCM
^ a ix•rH •H
43•U
co u o •iH
in
H
HHiCmGO
4-1
4->
•Hc
<0
4->
0)
434->
aCD
i—
i
(X• r-l
4J
(3•Ho
O -1-1
p-* >v >-i 4 )
C ,n r-l T3O o 0)
CD
434J
CD
in Q)
H uP
•H
cfl
XJa)
ecd
r,
•1-1
4-1•P J_l
i/j
125empera
43•i-l
t-t
cd
o
3cd
U
co
cu
b0cCO
4J
CO
Co
•I-l
4J
CD co
43 -H4J CO
CD
Eh4-1
a) u ^^O i—
i
CI)cO TS a>
o crt rjS-l
CD S-i
o CO > a 44 •H COH H •r-l4-i a*
ou a>
jirH
1—
1
CO•rH
4-> CD
u -iCJ C 43
CDi
)
Bl <u -uin
eo
cfl
1-1
4-)
c•H
(U
43HX)•rH CD
x)Sh a
pO CD «H
S-l CO
oin
co CDx>
S-l
cO C•-I
COo
SH 43•H4-4
3O5-1
CD S-i
> ?u u0) 4-1 a) 3 Sh
inCM
4-1
OCU
X!4-J
O <D
CD 4->
c S3 43 3o 5-1 4-1 nJ H O<H O O 434J l-l
•• 4->
o cfl r4 43 U • -r-l
50 fl) CJ G u $ccj
cd bU o
oSh
Q)
434J
0)
4->
S-l
cO
!-i co
Sh J->
<D C&J CO
C•H
4-> O,* C 0) 3 a,<J o bO o
<^ -U •H >^43 T3r-l
UJ 4-> X 4-> CD
CO O •H Xtu W 5-1 > -rH
!-i 0) 43 » 4-1
3 > •rH c CD
bO u 1—1 o X) CD
•-I 3 CO a) CO 43Ijh o CJ C E 4J
71
VI
u
uu &w P
0)
c e
•H•p eid 0>
U Aa h•Hr-i 0)
<o t)
o C10
<4-l pW•H
c wo IVH «•p10 str> p10 cft-HO PM (0
d. rHCM
o(Jl c1 -HV)H
o>
M3P(0
s-l
0)
ano»
oo
in uo> co> MM M Ho> ai pai 10
71 0) >-l
ai ac u •rH
H tr> ^h01 10
u doM M cM 0) Hw a
a> m-i -o • to » 1)
,£3 O C S-i 4-» ,0142H cd o C -P
}_| r-l >H CD
o • rj oti y• ^ E a) d cd cdM U 3 _ S„ 5 c y T) -o
£?w
.h 3 a> to cu
73.15
K
and
1234.
S
or
unit
negative
,
tin,
zinc,
alum
e
been
made
witho
and
the
three
fix
curves
for
error
mK
errors
incurr
CM S 5 W U rH, i) 3 « >h o •
c > «h * M u o0) «H "O (J) J-J
S .H •<-< W 4J*J W _. C CD M-l
<u o - 13 <u j-i oX! CX § g !-i Cd
3 S cd C
SPRTs unitgalli asst
the
pigure gatio
W<4-l cd
° x .. -o w o
of
calibration
lues
caused
by
te
range,
namel
point
of
water
ith
error
outsi
ncluded
in
thi
curves
show
pr
cd fl a) S ,H„,M > y H XH _ Q. 4-> O W
ors
from
erro
temperature
xed
points
in
on
at
the
tri]
he
fixed
poin
nthesis.
Als
of
water;
the
r 2 ,H >h *-> Q>U -^ <H +J . S-l 4-)
® J-> rt ^ cd C
ion
of
or
in
of
the
calibr
Lfied
b
the
p
le
poi
r.
ropagat
the
err
t
each
.
The
e
ident inside
he
trip
of
wate
a, co co j. u 4J
> *Jcd o 4J
g C C V-l 4-> f-• 2 O 'rH VI U Cd .Ho 'S "-i o cd <u oCM « 4J CX > U ex
cd j_4 -u o0) w >-i U 3 3 W a)J-i a> ,13 a) o o 3 rH
bO i-l i-l rH CD 4-> CD
ex•H
•H ^ Cd -H ^ -"I 43 S-l
fti O O W H £ 4J 4J
72
Q)
<T>
C(0
COCMCM
CO
HCOH
c
E-i <tu
T)
0)
H
o
a)
-P(0
a
1n
-p
w
in
inCM
0)
c(0
IT
oI
CM
oCO
o oI
in
oI
o oI
co
oI
oI
oHI
H O rH CM n f in vo r^ co Ch o• • • • • • • • • • o
O o O1
O1
ol
o1
U
Uu C
•H
oi
o1
oi
o1
o1 1
W <D4J i-H
•rH
•rH
W> 5-1
C 4->
in •H
CM
i-i
a)
X4J
O OCU 4->
cfl
5-1 eo10
4-> ObO
CM a; 5-1
>•>4-1
O
T) 4->
10 (D fi
CM o3
•Ho
•a CUo5-1 0)
4-> I-I
C a,•rH •rH
O 5-4
o 5-14->
CM O a)5-i XU •pQ)
a« Q> o .
in X4->£*
H aT8>SU
-P(0
M
W
OX
S o2 5
0) q) cOin
ft X °4J W3H o) 5-1
Eh 5-i nj
a)o > 4-1
O Oa)
X C 4->in u o cCNJ •rH .HH
ti oHP-i
cfl p,5-1
X <D
C/5•rH r-l
rj fxOoH
Heni—
i
Cfl .rH
o uau
1Z ? a>
nJ &$5-i
in o•H H
r^ 4-1
rH
5o
o - ?in S-i c 9O o ^
5-1 •rH5-1 4->W H
C ®
4-1 £<-l n
*CM O 4-1
•H O<u 4J5-1 cd 4->
s •H Cw > -H
•I-I <B OEn Tj &,
73
1c<o
u
in
mCN
CO
o
CT"
ini
eo ***
9%
o 800
osoc
(0
M•PX
CTic>
H
O inH in
ini
1 1 1 1 i i i i i i i i i i i i i i i i
/
/7
/
y
i i i i i i i i i i i i i i i i i i i
in4->
oc
CM b<w
Co1-1
4Jcd
o •r-l
in >C4 a>
in
CM
oo
in
oin
inCN
ooH
in
o inH
in inI
U KB
^ C0) -H
«
cu
3•Pm>-i
cu
scu
-iwC O
c•Hop.
4J
i-H
CO
a
CQ 4J
W)
U u
oO 5-1
co eu4-> IWX O
4->
c
<1)ex
o3 CO
TJ .-I
o a•u
C
a, |<u u
0)
O cfl
S M
4J CD
E-j d)
8* >P-i oC/>
Zccj
CO
° rj<4-l CO
> Bb o
00
O4-1
>^u
oM0)
a
<4-l
O
•U
c•1-1
oex
0>
r-H
IX•r-l
S-l
•P
CO
psou(M
T)
CO
S
o
<N O)
CM C
<l) 6 35-1 U >H3 0> r-l
t>0 +-> i—
I
•rH CO CO
En T-) no
74
1c
in
mw
O
in
oo
3
§8.
W
OS3
Is
o>o
"5u<
in
Hm
t
ooo
in
oi
~\1 r t 1 1 r
CM
OCM
m in
o
u «
U G
OO
in
oi
in
in
cm
ocm
ininCM
oinCM
inx*
cm ««.
o CD
*r 5-«
cm 3P(0
in 5-4
r> 0)
CM &sCD
o HnCM
inCMCM
oCM
+ mHCM
inoCM
CM
(0 CD
u x,•rH 4->
b0 OC 4J
«J J-l
O O
ctf CD
iJs
X! <4-4
co o
t^4->
•Ht) o .
a) as/o *,5 «> o"OhoO CXCN5-1
4J
c
CD
hi •£O 4->
5-1
n aco o
j-i
co cmX
5-1
o4-> CO >-•
bo CO
c sto
o
W CO
X4-> 4-1
Cfl
4->
CMCM
4->
C•Hoa
CO
CO
u
bO cfl
°SgH _. cw
P4 O T3C/3
'-,5-1
m CO 555 c o
V4 -CO g4J 3
£ -<3 r-lo<4-4
CO
>5-1
coCM
. cd
C b0O•H 14-J
4-> O
4-1
cd
4->
c•Ho
b0
•i-l 4J
CO
5-1
3 .
W) > i-l
•H CO CO
75
in
oi
HI
in
HI I
in
I
on
<0
tPC
uAto
in ua
in id
uPX
o wcr»
1
CO
HH<1>
rd
W
villi 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
k\\>V -
v\ -
\\\x\ \
-
\\-
\\\\
-
1 1 1 1 1 1 1 1 1 1 1 1
\ \ <M
\ \Mm \ \ro \ \
i \ t \ i
W S V•»\\ \ro \\
i \\ i i
3t
\ . 1 I I
inl
in
oCMi
in
OnI
inro
O<*
I
in
I
oinl
U
ou
•p(d
u
aea)
En
oo
in
o HI
in
i
egI
in
CMl
° e
a) .5
76
C V co
O C <D
0)
c(0
u
w
IT)
. ain
a)
o<r>
l
wH c
(d
M
3WH
o oCM
O•
OO
ot
HI
CMI
OCO
I
inl
in
oinco
oin
oinvo
oinin
oin
oinn
«
0)
u
-p
u0)
a.ea)
oinCM
o o oo H
I
u «o nuu cQ) -H
fMI I
oin
I
4J Otd ex
>0)
0)
a>
0) 5-1
4->
<u
*&&C 4J•H4-> Otd -U
-5 Ma, <»
a) oo P<
^^sS.5 *j
w <g
O ^n s
(1)'4~l
Isw
a)
IdU
H °£ CPM ow .3
r-lw
z5
o C
<u su 3CO i-l
•h td
CO
C oo *J
a> td
bO S
u °•u
c•1-1
oex
o ®
«-i 5< u
•§
^ to
o° 0)
O 4J
ct\ td
r-l
O <D
4-> HO -
H <D
O 4->
ex td
bO*C <H
Otd <"
a)
>
u
uo
a) S
T3
m to
cm C
NCD
0)
)-l
tw
0)
,e4J
6o
tw
-d
4->
c•r4
oex
bO C•-H 3
CX•HH4J
0)
bOG
td td
& u
to
o <u
C 4=•r-l 4J
NM
tw OO "4-1
77
4.4 THERMOCOUPLES (77 K TO 2400 K)
There are numerous letter- designated types of thermocouples. The Type S
thermocouple was the standard instrument of the IPTS-68(75) in the range from630.74 °C to 1064.43 °C, but it is not a standard instrument of the ITS-90.Customer thermocouples are calibrated at NIST by using a set of temperaturefixed points, by comparison with SPRTs, or by comparison with reference -standardthermocouples that have been calibrated either by comparison with an SPRT or a
radiation pyrometer, or through the use of fixed points. For details of thecalibration procedures and of the uncertainties involved, see NIST SP 250-35
[22] and NIST Monograph 175 [23] (or Monograph 125 [82]).
Usually, the calibration data for most types of thermocouples are analyzedrelative to reference tables, such as those given in NBS Monograph 125 [82].Monograph 125, of course, has reference tables for thermocouples based on theIPTS-68(75). This monograph has been revised and updated to give referencetables for all letter-designated thermocouples based on the ITS-90. The revisedversion of Monograph 125 is Monograph 175 [23] and it supersedes Monograph 125.
The electromotive -force -temperature data for a thermocouple calibrated on theIPTS-68(75) can be converted to an approximate ITS-90 calibration through theuse of the differences (t90
- t68 ) given in table 1 of this document and the S
values in mV/°C for the relevant thermocouple given in Monograph 175 or Monograph125. An example of this conversion is given in table 10. A typical calibrationreport is presented in appendix 3 (see sec. 6.3.8).
4.5 LIQUID- IN -GLASS THERMOMETERS
Liquid- in- glass thermometers have uncertainties of realization as small as
± 30 mK in the temperature range from °C to about 100 °C, but deteriorates atlower and higher temperatures. Liquid- in- glass (primarily, mercury- in- glass)thermometers are calibrated at NIST by comparison with SPRTs in liquid baths ofvarious kinds that cover different temperature ranges. For details of the
calibration procedures and of the uncertainties involved, see NIST SP 250-23
[105]. An example of a calibration report, based on the IPTS-68(75), of a
liquid- in-glass thermometer is given in appendix 3 (see sec. 6.3.9). Acalibration report for the same thermometer, with the IPTS-68(75) calibrationconverted to an approximate ITS-90 calibration through the use of (t90 - t68 )
differences given in table 1 is given also in appendix 3 (see sec. 6.3.10).
4.6 INDUSTRIAL PLATINUM RESISTANCE THERMOMETERS
Industrial platinum resistance thermometers (IPRTs) are designed primarily for
use in the temperature range from about 77 K (approximate liquid nitrogen boilingpoint) to 500 °C. Typically, the manufacturer of IPRTs quotes minimuminstabilities of the IPRTs at the ± 0.1 K level over this range of temperatures.Some IPRTs may be somewhat better than this but others may be considerably worse.
As seen from table 1, the maximum difference of (r90 - r68 ) below 500 C C is about0.08 K, and therefore the difference in temperature due to the change intemperature scales is within the instability of many IPRTs. Continued use ofthe IPTS-68(75) and of equations and standards [American Society for Testing andMaterials (ASTM) Standard, E 1137, and International Electrotechnical Commission
78
Table 10. Example of a conversion of calibration values of a type K thermocoupleon the IPTS-68 to an approximate calibration on the ITS -90
Calibration Values Calibration Valueson IPTS -68 on ITS -90
t68(°C)
emf68(mV)
S
(mV/°C)Monograph
125
^90" ^68(°C)
Table 1
A[-S-.(t90 -t88 )]
(mV)
^90
(°C)
emf90a
(mV)
0.0 0.000 0.0395 0.000 0.000 0.0 0.000
100.0 4.092 0.0414 -0.026 -0.001 100.0 4.093
200.0 8.130 0.0400 -0.040 -0.002 200.0 8.132
300.0 12.195 0.0415 -0.039 -0.002 300.0 12.197
400.0 16.383 0.0422 -0.048 -0.002 400.0 16.385
500.0 20.633 0.0426 -0.079 -0.003 500.0 20.636
600.0 24.904 0.0425 -0.115 -0.005 600.0 24.909
700.0 29.136 0.0419 0.20 0.008 700.0 29.128
800.0 33.288 0.0410 0.34 0.014 800.0 33.274
900.0 37.338 0.0400 -0.01 0.000 900.0 37.338
1000.0 41.281 0.0389 -0.19 -0.007 1000.0 41.288
1100.0 45.118 0.0378 -0.26 -0.010 1100.0 45.128
emf,90 emf6a- A
(IEC) Standard, Publication 751] referenced to the IPTS-68(75), therefore, wouldresult in an increase in uncertainty of temperature of only about 0.1 K if the
temperature were expressed as being on the ITS -90. (Note: the ASTM and the IEC
are converting their respective IPRT tables from the IPTS -68 (75) to the ITS -90.
ASTM Committee E-20 on Temperature Measurements is responsible for thisconversion for the ASTM)
.
When IPRTs are calibrated on the ITS -90, of course, they may be calibrated in
the same manner as is used for SPRTs . A better method of calibrating IPRTs,however, is to obtain resistance -temperature data by comparison with a calibratedSPRT at numerous temperatures over the range of interest and then fit a
polynomial in t90 to £(t90 ), to R(t90 )/R(0 °C) , or to R(t90 )/R(0 . 01 °C) data by a
least squares technique.
4.7 THERMISTOR THERMOMETERS. DIGITAL THERMOMETERS. AND OTHER TYPES OFTHERMOMETERS
Thermistor thermometers, digital thermometers (with resistance, thermocouple,or diode sensors) , and other types of thermometers are calibrated at NIST bycomparison with SPRTs in liquid baths. The calibration procedures followed aresimilar to those used with liquid- in-glass thermometers. The uncertainties ofcalibration range from as small as ± 2 mK for thermistor thermometers to tenths
79
of kelvins for the others . The temperatures of calibration for these types ofthermometers usually lie somewhere within the range from about 77 K to 850 K.
Bead- in- glass probe type thermistors used in the moderate temperature range arequite stable and they may be used to approximate the ITS -90 at a level of about+ 1.5 mK to ± 2.0 mK [67]. In their case, a polynomial, the degree of whichdepends on the temperature range of the calibration, is fitted to resistance-temperature data and the results reported in terms of that polynomial. Acalibration on the IPTS-68(75) may be converted to an approximate ITS -90
calibration by the same procedure as outlined for RIRTs
.
4.8 THE LOGO OF THE NATIONAL CONFERENCE OF STANDARDS LABORATORIES FOR THE ITS -90
The National Conference of Standards Laboratories (NCSL) formed an Ad HocCommittee on the Change of the Temperature Scale at the beginning of 1988 inorder to publicize the new temperature scale (ITS-90) and to facilitate its
implementation. At the NCSL meeting in July 1989, the Ad Hoc Committee adopteda logo, available from the NCSL [1800 30th Street, Suite 305B, Boulder, CO 80301,Tel. (303) 440-3339], that may be affixed to thermometers that have beencalibrated on the ITS-90. The purpose of the logo, illustrated in figure 26,
is to indicate at a glance, without having to refer to documentation, thosethermometers in a laboratory that have been calibrated on the new scale. Thisis particularly useful for those laboratories that have their variousthermometers calibrated on a prescribed schedule, with different thermometersbeing calibrated at different times.
80
ITS-90
Figure 26. The NCSL ITS-90 logo
81
5 . REFERENCES
[I] Ancsin, J., Vapour Pressures and Triple Point of Neon and the Influenceof Impurities on These Properties, Metrologia 14, 1-7 (1978).
[2] Ancsin, J., Thermometric Fixed Points of Hydrogen, Metrologia 13, 79-86
(1977).
[3] Ancsin, J., A Study of the Realization of the Melting and Freezing Pointsof Silver, Metrologia 26, 167-174 (1989).
[4] Ancsin, J., Melting Curves and Heat of Fusion of Indium, Metrologia 21 .
7-9 (1985).
[5] Barber, C. R. , Handley, R. , and Herington, E. F. G. , The Preparation andUse of Cells for the Realization of the Triple Point of Water, Brit. J. Appl.Phys. 5, 41-44 (1954).
[6] Barber, C. R. , A Proposal for a Practical Scale of Temperature Below 20 K,
Temperature . Its Measurement and Control in Science and Industry . Edited by H.
H. Plumb, Vol. 4, Part 1, pp. 99-103 (Instrument Society of America, Pittsburgh,1972).
[7] Barrick, P. L. , Brown, L. F. , Hutchinson, H. L. , and Cruse, R. L. , ImprovedFerric Oxide Gel Catalysts for Ortho-Parahydrogen Conversion, Edited by K. D.
Timmerhaus, Vol. 10, paper D-l, 131-189 (Plenum Press, New York, NY, 1965).
[8] Bauer, G., and Bischoff, K. , Evaluation of the Emissivity of a CavitySource by Reflection Measurements, Applied Optics 10, 2639-2643 (1971).
[9] Bedford, R. E. , Effective Emissivities of Blackbody Cavities -- A Review,Temperature. Its Measurement and Control in Science and Industry . Edited by H.
H. Plumb, Vol. 4, Part 1, pp. 425-434 (Instrument Society of America, Pittsburgh,PA, 1972).
[10] Bedford, R. E. and Ma, C. K. , Emissivities of Diffuse Cavities, II:
Isothermal and Nonisothermal Cyl indro- cones , J. Opt. Soc. Am. 65., 565-572 (1974).
[II] Bedford, R. E. and Ma, C. K. , Emissivities of Diffuse Cavities: Isothermaland Nonisothermal Cones and Cylinders, J. Opt. Soc. Am. 64, 339-349 (1974).
[12] Belecki, N. B. , Dziuba, R. F. , Field, B. F. , and Taylor, B. N.
,
Guidelines for Implementing the New Representations of the Volt and Ohm EffectiveJanuary 1, 1990, NIST Technical Note 1263 (June 1989).
[13] Berry, K. H. , NPL-75: A Low Temperature Gas Thermometry Scale from 2.6 Kto 27.1 K, Metrologia 15, 89-115 (1979).
[14] BIPM Com. Cons. Thermometrie, 17, 1989, in press.
[15] Bongiovanni, G. , Crovini, L. , and Marcarino P. , Effects of Dissolved Oxygenand Freezing Techniques on the Silver Point, Metrologia 11 . 125-132 (1975).
82
[16] Bonhoure, J. and Pello, R. , Temperature du Point Triple du Gallium,Metrologia 19, 15-20 (1983).
[17] Bonhoure, J. and Pello, R. , Points Triples de l'Argon et du Methane:Utilisation de Cellules Scellees , Metrologia 16, 95-99 (1980).
[18] Bonhoure, J. and Pello, R. , Temperature of the Triple Point of Methane,Metrologia 14, 175-177 (1978).
[19] Bonhoure, J. and Terrien, J., The New Manobarometer of the BureauInternational des Poids et Mesures, Metrologia 4, 59-68 (1968).
[20] Bonnier, G. and Hermier, Y. , Thermal behavior of thermometric sealed cellsand of a multi- compartment cell, Temperature . Its Measurement and Control inScience and Industry . Edited by J. F. Schooley, Vol. 5, Part 1, pp. 231-237(American Institute of Physics, New York, 1982).
[21] Bonnier, G. and Moser, A., Development at the Institut National de
Metrologie of sealed cells as IPTS fixed points, Measurement (IMEKO) 1, 143-151
(1983).
[22] Burns, G. W. and Scroger, M. G. , The Calibration of Thermocouples andThermocouple Materials, NIST Special Publication 250-35 (April 1989).
[23] Burns, G. W. and Scroger, M. G., Temperature -electromotive Force ReferenceFunctions and Tables for Letter -designated Thermocouple Types Based on theITS-90, NIST Monograph 175 (1990).
[24] Cataland, G., Edlow, M. H. , and Plumb, H. H., Recent Experiments on LiquidHelium Vapor Pressure Measurements from 2 °K to 4 °K, Temperature . Its
Measurement and Control in Science and Industry . Edited by F. G. Brickwedde, Vol.
3, Part 1, pp. 413-417 (Reinhold Publishing Corporation, New York, 1962).
[25] Chappuis, P., Etudes sur le Thermometre a Gaz et Comparison des
Thermometres a Mercure avec le Thermometre a Gaz, Trav. et Mem. BIPM 6, 1-125
plus 187 pages of observations (1888)
.
[26] Chattle, M. V., Rusby, R. L. , Bonnier, G. , Moser, A., Renaot, E.,
Marcarino, P. , Bongiovanni, G. , and Frassineti, G. , An intercomparison of galliumfixed point cells, Temperature . Its Measurement and Control in Science andIndustry . Edited by J . F. Schooley, Vol. 5, Part 1, pp. 311-316 (AmericanInstitute of Physics, New York, 1982).
[27] Coates , P. B. and Andrews , J. W. , A precise determination of the freezingpoint of copper, J. Phys . F: Metal Phys
., 8, 277-285 (1978).
[28] Coates, P. B. , The NPL photon- counting pyrometer, Temperature Measurement.1975 . The Institute of Physics Conference Series Number 26, Edited by B. F.
Billing and T. J. Quinn, Chapter 5, pp. 238-243 (The Institute of Physics,London, 1975).
83
[29] Comptes Rendus des Seances de la Dix-huitieme Conference Generale des
Poids et Mesures, Resolution 7, p. 101 (1987).
[30] Comptes Rendus des Seances de la Dixieme Conference Generale des Polds etMesures, Resolution 3, p. 79 (1954).
[31] Compton, J. P., The Realization of Low Temperature Fixed Points,Temperature . Its Measurement and Control in Science and Industry . Edited by H.
H. Plumb, Vol. 4, Part 1, pp. 195-209 (Instrument Society of America, Pittsburgh,PA, 1972).
[32] Compton, J. P. and Ward, S. D. , Realization of the Boiling and TriplePoints of Oxygen, Metrologia 12, 101-113 (1976).
[33] Darling, A. S. and Selman, G. L. , Some Effects of Environment on thePerformance of Noble Metal Thermometers, Temperature . Its Measurement andControl in Science and Industry . Edited by H. H. Plumb, Vol. 4, Part 3, pp. 1633-
1644 (Instrument Society of America, Pittsburgh, PA, 1972).
[34] DeWitt, D. P. and Nutter, G. D. , Theory and Practice of RadiationThermometry (John Wiley and Sons, Inc., New York, NY, 1988).
[35] DeWitt, D. P. and Hernicz , R. S., Theory and Measurement of EmittanceProperties for Radiation Thermometry Applications , Temperature. Its Measurementand Control in Science and Industry . Edited by H. H. Plumb, Vol. 4, Part 1, pp.459-482 (Instrument Society of America, Pittsburgh, PA, 1972).
[36] Ditmars, D. A. and G. T. Furukawa, Detection and Damping of Thermal-acoustic Oscillations in Low- temperature Measurements , J. Res. Natl. Bur. Stands.69C, 35-38 (1965).
[37] Durieux, M. and Rusby, R. L. , Helium Vapour Pressure Equations on theEPT-76, Metrologia 19, 67-72, (1983).
[38] Durieux, M. , van Dijk, J. E. , ter Harmsel, H. , Rem, P. C, and Rusby, R.
L. , Helium Vapor Pressure Equations on the EPT-76, Temperature . Its Measurementand Control in Science and Industry . Edited by J . F. Schooley, Vol. 5, Part 1,
pp. 145-154 (American Institute of Physics, New York, 1982).
[39] Elliott, R. P., Constitution of Binary Alloys, First Supplement (McGraw-
Hill Book Company, New York, NY, 1965).
[40] Ferguson, J. A., Realization of the Triple Point of Water, J. Phys . E: Sci.
Instrum. 3, 447-451 (1970).
[41] Furukawa, G. T. , Reproducibility of the Triple Point of Argon in SealedTransportable Cells, Temperature. Its Measurement and Control in Science andIndustry . Edited by J . F. Schooley, Vol. 5, Part 1, pp. 239-248 (AmericanInstitute of Physics, New York, NY, 1982).
84
[42] Furukawa, G. T. , Bigge, W. R. , and Riddle, J. L. , Triple Point of Argon,Temperature, Its Measurement and Control in Science and Industry . Edited by H.
H. Plumb, Vol. 4, Part 1, pp. 231-243 (Instrument Society of America, Pittsburgh,PA, 1972).
[43] Furukawa, G. T. , Riddle, J. L. , Bigge. W. R. , and Pfeiffer, E. R.,
Application of Some Metal SRM's as Thermometric Fixed Points, NBS SpecialPublication 260-77, 140 pages, August 1982.
[44] Furukawa, G. T. , Piccirelli, J. H. , Reilly, M. L. , Cryoscopic Determinationof the Purity of Benzene by Calorimetry, Purity Determinations by the ThermalMethods . Edited by R. L. Blaine and C. K. Schoff, pp. 90-106 (American Societyfor Testing and Materials, Philadelphia, PA, 1984).
[45] Furukawa, G. T. , Investigation of Freezing Temperatures of National Bureauof Standards Aluminum Standards, J. Res. Natl. Bur. Stand. (U.S.) 78A, 477-495
(1974).
[46] Furukawa, G. T. , Riddle, R. L. , and Bigge, W. R. , Investigation of freezingtemperatures of National Bureau of Standards tin standards, Temperature . Its
Measurement and Control in Science and Industry . Edited by H. H. Plumb, Vol. 4,
Part 1, pp. 247-263 (Instrument Society of America, Pittsburgh, PA, 1972).
[47] Furukawa, G. T., The Triple Point of Oxygen in Sealed Transportable Cells,
J. Res. Natl. Bur. Stands. (U.S.) 91, 255-275 (1986).
[48] Gordon, C. L. and Wichers , E. , Purification of Mercury and Its PhysicalProperties, Ann. New York Acad. Sci. 65, 369-387 (1957).
[49] Guildner, L. A. and Edsinger, R. E. , Deviation of International PracticalTemperatures from Thermodynamic Temperatures in the Temperature Range from273.16 K to 730 K, J. Res. Bur. Stands. 80A, 703-738 (1976).
[50] Guildner, L. A. , Stimson, H. F. , Edsinger, R. E. , and Anderson, R. L. , AnAccurate Mercury Manometer for NBS Gas Thermometer, Metrologia 6, 1-18 (1970).
[51] Hansen, M. , Constitution of Binary Alloys (McGraw-Hill Book Company, Inc.
,
New York, NY, 1958).
[52] Harrison, E. R. , Hatt, D. J., Prowse, D. B. , and Wilbur-Ham, J., A NewInterferometric Manometer, Metrologia 12, 115-122 (1976).
[53] Jangg, G. and Palman, H. , Die Loslichkeit verschiedener Metalle in
Quecksilber, Zeit. Metallkde. 54, 364-369 (1963).
[54] Jones, T. P. and Tapping, J., A photoelectric pyrometer temperature scalebelow 1064.43 C
C and its use to measure the silver point, Temperature . Its
Measurement and Control in Science and Industry . Edited by J. F. Schooley, Vol.
5, Part 1, pp. 169-174 (American Institute of Physics, New York, 1982).
85
[55] Jones, T. P. and Tapping, J., The Determination of the ThermodynamicTemperatures of Thermometry Fixed Points in the Range 660 °C to 1064 °C,
Metrologia 25, 41-47 (1988).
[56] Jung, J. J., Determination of the difference between the thermodynamicfixed-point temperatures of gold and silver by radiation thermometry, TemperatureMeasurement. 1975 . The Institute of Physics Conference Series Number 26, Editedby B. F. Billing and T. J. Quinn, Chapter 5, pp. 278-286 (The Institute ofPhysics, London, 1975).
[57] Jung, H. J. , A Measurement of Thermodynamic Temperatures Between 683 K and933 K by an Infrared Pyrometer, Metrologia 23, 19-31 (1986).
[58] Kemp, R. C. and Kemp, W. R. G. , The Triple Point, Boiling Point and 17 KPoint of Equilibrium Hydrogen, Metrologia 15, 155-159 (1979).
[59] Kemp, R. C, Kemp, W. R. C, and Cowan, J. A., The Boiling Points andTriple Points of Oxygen and Argon, Metrologia 12, 93-100 (1976).
[60] Kemp, R. C, Kemp, W. R. G. , and Besley, L. M. , A Determination ofThermodynamic Temperatures and Measurements of the Second Virial Coefficient ofAHe Between 13.81 K and 287 K Using a Constant -Volume Gas Thermometer, Metrologia23, 61-86 (1986/87)
[61] Klein, H. -H. , Klempt, G. , and Storm, L. , Measurement of the ThermodynamicTemperature of AHe at Various Vapour Pressures by a Noise Thermometer, Metrologia15, 143-154 (1979).
[62] Lee, R. D. , Construction and Operation of a Simple High-Precision Copper-Point Blackbody and Furnace, NBS Technical Note 483, May 1969.
[63] Maghenzani, R. , Molinar, G. F. , Marzola, L. , and Kulshrestha, R. K. ,
Pressure Metrology up to 5 MPa in Different Gas Media, J. Phys . E: Sci. Instrum.
20, 1173-1179 (1987).
[64] Mangum, B. W. , Platinum Resistance Thermometer Calibrations, NBS SpecialPublication 250-22 (October 1987).
[65] Mangum, B. W. , Triple point of gallium as a temperature fixed point,Temperature . Its Measurement and Control in Science and Industry . Edited by J.
F. Schooley, Vol. 5, Part 1, pp. 299-309 (American Institute of Physics, NewYork, 1982).
[66] Mangum, B. W. , Special Report on the International Temperature Scale of1990; Report on the 17th Session of the Consultative Committee on
Thermometry, J. Res. Natl. Inst. Stand. Technol. 95, 69-77 (1990).
[67] Mangum, B. W. , Triple point of succinonitrile and its use in the
calibration of thermistor thermometers, Rev. Sci. Instrum. 54, 1687-1692 (1983).
[68] Mangum, B. W. and Thornton, D. D. , Determination of the Triple-PointTemperature of Gallium, Metrologia 15, 201-215 (1979).
86
[69] Mangum, B. W. Determination of the Indium Freezing-point and Triple-pointTemperatures, Metrologia 26, 211-217 (1989).
[70] McAllan, J. V. and Ammar, M. M. , Comparison of the freezing points ofaluminum and antimony, Temperature , Its Measurement and Control in Science andIndustry . Edited by H. H. Plumb, Vol. 4, Part 1, pp. 273-285 (Instrument Societyof America, Pittsburgh, PA, 1972).
[71] McConville, G. T. , The Effect of the Measuring Tube Surface onThermomolecular Pressure Corrections in Vapor Pressure Thermometry,Temperature. Its Measurement and Control in Science and Industry . Edited by H.
H. Plumb, Vol. 4, Part 1, pp. 159-165 (Instrument Society of America, Pittsburgh,PA, 1972).
[72] McLachlan, A. D., Uchiyama, H. , Saino, T. , and Nakaya, S., The Stabilityof the Freezing Point of Copper as a Temperature Standard, Temperature . ItsMeasurement and Control in Science and Industry . Edited by H. H. Plumb, Vol. 4,
Part 1, pp. 287-293 (Instrument Society of America, Pittsburgh, PA, 1972).
[73] McLaren, E. H. , The Freezing Points of High Purity Metals as PrecisionTemperature Standards. II. An Investigation of the Freezing Temperatures ofZinc, Cadmium, and Tin, Can. J. Phys . 35, 1086-1106 (1957).
[74] McLaren, E. H. , The Freezing Points of High Purity Metals as PrecisionTemperature Standards. IV. Indium: Thermal Analyses on Three Grades ofCadmium, Can. J. Phys. 36, 1131-1147 (1958).
[75] McLaren, E. H. , The Freezing Points of High Purity Metals as PrecisionTemperature Standards. III. Thermal Analysis on Eight Grades of Zinc withPurities Greater than 99.99+%, Can. J. Phys. 36. 585-598 (1958).
[76] McLaren, E. H. and Murdock, E. G., The Freezing Points of High PurityMetals as Precision Temperature Standards. V. Thermal Analyses on 10 Samplesof Tin with Purities Greater than 99.99+%, Can. J. Phys. 38, 100-118 (1960).
[77] Ohtsuka, M. and Bedford, R. E. , Measurement of the thermodynamictemperature interval between the freezing points of silver and copper,Temperature . Its Measurement and Control in Science and Industry . Edited by J.
F. Schooley, Vol. 5, Part 1, pp. 175-181 (American Institute of Physics, NewYork, 1982).
[78] Orlova, M. P., Astrov, D. N. , Sharevskaya, D. I., Belyanskii, L. B.,
Razhba, Ya. E. , and Khnykov, V. M. , Primary State Standard for the Unit ofTemperature in the Range of 13.81-273.15 K, Measurement Techniques 16, 477-482
(1973).
[79] Pavese, F. , Ancsin, J., Astrov, D. N. , Bonhoure , J., Bonnier, G. , Furukawa,
G. T. , Kemp, R. C, Maas , H. , Rusby, R. L. , Sakurai, H. , and Ling Shan-Kang, AnInternational Intercomparison of Fixed Points by Means of Sealed Cells in the
Range 13.81 K - 90.686 K, Metrologia 20, 127-144 (1984).
87
[80] Pavese.F., The Triple Points of Argon and Oxygen, Metrologia 14, 93-103,
(1978).
[81] Peggs, G. N. , Elliott, K. W. T. , and Lewis, S., An Intercomparison Betweena Primary Standard Mercury Barometer and a Gas -operated Pressure BalanceStandard, Metrologia 15, 77-85 (1979).
[82] Powell, R. L. , Hall, W. J., Hyink, C. H. Jr., Sparks, L. L. , Burns, G. W.,
Scroger, M. G. , and Plumb, H. H. , Thermocouple Reference Tables Based on theIPTS-68, NBS Monograph 125 (March 1974).
[83] Preston- Thomas, H. , The International Temperature Scale of 1990 (ITS-90),Metrologia 27, 3-10 (1990).
[84] Proces-Verbaux des Seances du Comite International des Poids et Mesures,
(78e session, octobre 1989), in press.
[85] Quinn, T. J., News from the BIPM, Metrologia 26, 69-74 (1989).
[86] Quinn, T. J. and Chandler, T. R. D. , The Freezing Point of PlatinumDetermined by the NPL Photoelectric Pyrometer, Temperature . Its Measurement andControl in Science and Industry , Edited by H. H. Plumb, Vol. 4, Part 1, pp. 295-
309 (Instrument Society of America, Pittsburgh, PA, 1972).
[87] Quinn, T. J., Temperature (Academic Press, Inc., New York, N. Y. (1983).
[88] Ricci, J. E. , The Phase Rule and Heterogeneous Equilibrium, (DoverPublications, Inc., New York, NY, 1966).
[89] Righini, F. , Rosso, A., and Ruffino, G., Temperature Dependence ofEffective Wavelength in Optical Pyrometry, Temperature. Its Measurement andControl in Science and Industry . Edited by H. H. Plumb, Vol. 4, Part 1, pp. 413-
421 (Instrument Society of America, Pittsburgh, PA, 1972).
[90] Roberts, T. R. , Sherman, R. H. , and Sydoriak, S. G. , The 1962 3He Scale ofTemperatures III. Evaluation and Status, J. Res. Natl. Bur. Stands. 68A . 567-578
(1964).
[91] Rusby, R. L. and Swenson, C. A., A New Determination of the Helium VapourPressure Scales Using a CMN Magnetic Thermometer and the NPL- 75 Gas ThermometerScale, Metrologia 16, 73-87 (1980).
[92] Sawada, S., Realization of the triple point of Indium in a sealed glasscell, Temperature . Its Measurement and Control in Science and Industry . Editedby J . F. Schooley, Vol. 5, Part 1, pp. 343-346 (American Institute of Physics,New York, 1982)
.
[93] Simon, M. , On the Phase Separation in the Liquid System Ne-pH2 , Phys
.
Letters, 5, 319 (1963).
[94] Sostman, H. E. , Melting point of gallium as a temperature calibrationstandard, Rev. Sci. Instrum. 48, 127-130 (1977).
88
[95] Sparrow, E. M. , Albers , L. U. , and Eckert, E. R. G. , Thermal RadiationCharacteristics of Cylindrical Enclosures, J. Heat Transfer C84 . 73-81 (1962).
[96] Steur, P. P. M. and Durieux, M. , Constant -Volume Gas Thermometry Between4 K and 100 K, Metrologia 23, 1-18 (1986).
[97] Streett, W. B. and D. H. Jones, Liquid Phase Separation and Liquid-vaporEquilibrium in the System Neon- Hydrogen, J. Chem. Phys . 42, 3989-3994 (1965).
[98] Takiya, M. , Precise Measurement of the Freezing Point of Silver with a
Platinum Resistance Thermometer, Comitee Consultatif de Thermometrie. 12thSession, Annexe T30, T154-T159 (1978).
[99] The 1976 Provisional 0.5 K to 30 K Temperature Scale, Metrologia 15, 65-68
(1979).
[100] The International Practical Temperature Scale of 1968, Metrologia 5, 35-44
(1969).
[101] The International Practical Temperature Scale of 1968, Amended Edition of1975, Metrologia 12, 7-17 (1976).
[102] Weber, S. and Schmidt, G. , Experimentelle Untersuchungen viber dieThermomolekulare Druckdifferenz in der Nahe der Grenzbedingung P]/p2
=(
r
£ 1/T2 )1/2
und Vergleichung mit der Theorie, Leiden Communication 246C, 1-13 (1936).
[103] Weeks, J. R. , Liquidus Curves and Corrosion of Fe, Cr, Ni, Co, V, Cb , Ta,
Ti, Zr, in 500-750 °C Mercury, Corrosion 23, 98-106 (1967).
[104] Weitzel, D. H. and Park, 0. E. , Iron Catalyst for Production of Liquidpara -Hydrogen, Rev. Sci. Instr. 27,57-58 (1956).
[105] Wise, Jacquelyn, Liquid- In-Glass Thermometer Calibration Service, NISTSpecial Publication 250-23 (September 1988).
89
APPENDICES
In these appendices, we present information on or copies of some of the relevantdocuments or meetings concerning the ITS -90, typical NIST calibration reportsof thermometers, and membership of the National Conference of StandardsLaboratories (NCSL) Ad Hoc Committee on the Change of the Temperature Scale
6.1 In this appendix, copies of, or excerpts from, some papers related to thework of the international bodies concerned with the ITS -90 are presented.
6.1.1 Reproduction of the article by H. Preston-Thomas, entitled, "TheInternational Temperature Scale of 1990 (ITS-90)," published in Metrologia 27,3-10 (1990), with corrections . Reprinted with permission of the BIPM.
90
Metrologia 27, 3-10(1990) metrologia© Springer-Verlag 1990
This copy incorporates textual corrections detailed in Metrologia 27,
107 (1990)
The International Temperature Scale of 1990 (TTS-90)
H. Preston-Thomas
President of the Comite Consultatif de Thermometrie and Vice-President of the Comite International des Poids et Mesures
Division of Physics, National Research Council of Canada, Ottawa, K1A0S1 Canada
Received: October 24, 1989
Introductory Note
The official French text of the ITS-90 is published by the
BIPM as part of the Proces-verbaux of the Comite Inter-
national des Poids et Mesures (CIPM). However, the
English version of the text reproduced here has been au-
thorized by the Comite Consultatif de Thermometrie
(CCT) and approved by the CIPM.
The International Temperature Scale of 1990
The International Temperature Scale of 1990 was adopt-
ed by the International Committee of Weights and Mea-sures at its meeting in 1989, in accordance with the re-
quest embodied in Resolution 7 of the 18th General
Conference of Weights and Measures of 1987. This scale
supersedes the International Practical Temperature Scale
of 1968 (amended edition of 1975) and the 1976 Provi-
sional 0,5 K to 30 K Temperature Scale.
1. Units of Temperature
The unit of the fundamental physical quantity known as
thermodynamic temperature, symbol T, is the kelvin,
symbol K, defined as the fraction 1/273,16 of the thermo-
dynamic temperature of the triple point of water '.
Because of the way earlier temperature scales were
defined, it remains common practice to express a temper-
ature in terms of its difference from 273,15 K, the ice
point. A thermodynamic temperature, T, expressed in this
way is known as a Celsius temperature, symbol t, defined
by:
t/°C=T/K-273,15. (1)
The unit of Celsius temperature is the degree Celsius,
symbol °C, which is by definition equal in magnitude to
the kelvin. A difference of temperature may be expressed
in kelvins or degrees Celsius.
The International Temperature Scale of 1990 (ITS-90)
defines both International Kelvin Temperatures, symbol
T90 , and International Celsius Temperatures, symbol f90 .
The relation between T90 and t90 is the same as that
between Tand t, i.e.:
t90/°C = T90/K-273,15 (2)
1 Comptes Rendus des Seances de la Treizieme Conference Gene-
rale des Poids et Mesures (1967- 1968), Resolutions 3 and 4. p. 104
The unit of the physical quantity Tgo is the kelvin,
symbol K, and the unit of the physical quantity r90 is the
degree Celsius, symbol °C, as is the case for the thermody-
namic temperature T and the Celsius temperature t.
2. Principles of the International Temperature Scale
of 1990 (ITS-90)
The ITS-90 extends upwards from 0,65 K to the highest
temperature practicably measurable in terms of the
Planck radiation law using monochromatic radiation.
The ITS-90 comprises a number of ranges and sub-ranges
throughout each of which temperatures T90 are defined.
Several of these ranges or sub-ranges overlap, and where
such overlapping occurs, differing definitions of T90 exist:
these differing definitions have equal status. For measure-
ments of the very highest precision there may be detect-
able numerical differences between measurements madeat the same temperature but in accordance with differing
definitions. Similarly, even using one definition, at a tem-
perature between defining fixed points two acceptable
interpolating instruments (e.g. resistance thermometers)
may give detectably differing numerical values of T90 . In
virtually all cases these differences are of negligible practi-
cal importance and are at the minimum level consistent
with a scale of no more than reasonable complexity: for
further information on this point, see "Supplementary
Information for the ITS-90" (BIPM-1990).
Reprinted with permission of BIPM,
J i L-200 200 400 600 800 1000
Fig. 1. The differences (t9Q — t6S ) as a
function of Celsius temperature t90
The ITS-90 has been constructed in such a way that,
throughout its range, for any given temperature the nu-
merical value of T90 is a close approximation to the nu-
merical value of T according to best estimates at the time
the scale was adopted. By comparison with direct mea-surements of thermodynamic temperatures, measure-
ments of T90 are more easily made, are more precise andare highly reproducible.
There are significant numerical differences between
the values of T90 and the corresponding values of T6Smeasured on the International Practical TemperatureScale of 1968 (IPTS-68), see Fig. 1 and Table 6. Similarly
there were differences between the IPTS-68 and the Inter-
national Practical Temperature Scale of 1948 (IPTS-48),
and between the International Temperature Scale of 1948
(ITS-48) and the International Temperature Scale of 1927
(ITS-27). See the Appendix and, for more detailed infor-
mation, "Supplementary Information for the ITS-90".
3. Definition of the International Temperature Scale
of 1990
Above the freezing point of silver (961,78CC) T90 is
defined in terms of a defining fixed point and the Planck
radiation law.
The defining fixed points of the ITS-90 are listed in
Table 1 . The effects of pressure, arising from significant
depths of immersion of the sensor or from other causes,
on the temperature of most of these points are given in
Table 2.
3.1. From 0,65 K to 5.0 K: Helium Vapour-Pressure
Temperature Equations
In this range T90 is defined in terms of the vapour pressure
p of3He and 4He using equations of the form:
T90/K = A + £ A-, [(In ( p/Pa) - B)/C][
. (3)
The values of the constants A , A { , B and C are given in
Table 3 for3He in the range of 0,65 K to 3,2 K, and for
4He in the ranges 1,25 K to 2,1768 K (the A point) and2,1768 K to 5,0 K.
Between 0,65 K and 5.0 K T90 is defined in terms of the
vapour-pressure temperature relations of3He and 4
He.
Between 3,0 K and the triple point of neon(24,5561 K) T90 is defined by means of a helium gas ther-
mometer calibrated at three experimentally realizable
temperatures having assigned numerical values (defining
fixed points) and using specified interpolation procedures.
Between the triple point of equilibrium hydrogen
(13,8033 K)and the freezing point of silver (961,78 °C) T90is defined by means of platinum resistance thermometers
calibrated at specified sets of defining fixed points andusing specified interpolation procedures.
3.2. From 3,0 K to the Triple Point ofNeon (24,5561 K)
:
Gas Thermometer
In this range Tgo is defined in terms of a 3He or a4He gas
thermometer of the constant-volume type that has been
calibrated at three temperatures. These are the triple
point of neon (24,5561 K), the triple point of equilibrium
hydrogen (13,8033 K), and a temperature between 3,0 Kand 5,0 K. This last temperature is determined using a3He or a
4He vapour pressure thermometer as specified in
Sect. 3,1.
Table 1. Defining fixed points of the ITS-90
Num- Temperature
ber
Sub- State" Wr(T90 )
stance3
T90/K t,o/°C
5
6
7
8
9
10
11
12
13
14
15
16
17
3 to 5
13,8033
wl7 :
•270,15 to
-268,15
-259,3467
-256,15
*20,3 a -252,85
He V
e-H 2 Te-H 2
V(or He) (or G)e-H 2 V(or He) (or G)
24,5561
54,3584
83,8058
234,3156
273,16
302,9146
429,7485
505,078
692,677
933,473
1234,93
1337,33
1357,77
-248,5939
-218,7916
-189,3442
-38,8344
0,01
29,7646
156,5985
231,928
419,527
660,323
961,78
1064,18
1084,62
Neo 2
ArHg
H 2
GaIn
Sn
ZnAl
AgAuCu
TTTT
TMFF
FFFFF
0,001 190 07
0,008 449 74
0,091 718 04
0,215 859 75
0,844 142 11
1,000 000 00
1,118 138 89
1,609 801 85
1,892 797 68
2,568 917 30
3,376 008 60
4,286 420 53
' All substances except 3He are of natural isotopic composition,
e-H2
is hydrogen at the equilibrium concentration of the ortho- andpara-molecular formsb For complete definitions and advice on the realization of these
various states, see "Supplementary Information for the ITS-90". Thesymbols have the following meanings: V: vapour pressure point;
T: triple point (temperature at which the solid, liquid and vapour
phases are in equilibrium); G: gas thermometer point; M, F: melting
point, freezing point (temperature, at a pressure of 101 325 Pa, at
which the solid and liquid phases are in equilibrium)
Table 2. Effect of pressure on the temperatures of some defining
fixed points*
Substance Assigned Temperature Variation
value of with pressure, p with depth, /
equilibrium (dT/dp)/ (dT/d/)/
temperature (10" 8 KPa" 1 )* (10~ 3 K- m -1)**
T90/K
e-Hydrogen (T) 13,8033 34 0,25
Neon (T) 24,5561 16 1,9
Oxygen (T) 54,3584 12 1,5
Argon (T) 83,8058 25 3,3
Mercury (T) 234.3156 5,4 7,1
Water (T) 273,16 -7,5 -0,73
Gallium 302,9146 -2,0 -1.2
Indium 429,7485 4,9 3,3
Tin 505,078 3,3 2 2
Zinc 692,677 4.3 2,7
Aluminium 933,473 7,0 1,6
Silver 1234,93 6,0 5,4
Gold 1337.33 6,1 10
Copper 1357,77 3.3 2,6
* Equivalent to millikelvins per standard atmosphere** Equivalent to millikelvins per metre of liquid
* The Reference pressure for melting and freezing points is the
standard atmosphere (p =101 325 Pa). For triple points (T) the
pressure effect is a consequence only of the hydrostatic head of
liquid in the cell
Table 3. Values of the constants for the helium vapour pressure
Eqs. (3), and the temperature range for which each equation, identi-
fied by its set of constants, is valid
3He 4He 4He0,65 K to 3,2 K 1,25 K to 2,1768 K 2,1768 K to 5,0 K
AA,
*2
A,
A,
KAnA,
A 9
BC
1,053 447
0,980 106
0,676 380
0,372 692
0,151 656
-0,002 263
0,006 596
0,088 966
-0,004 770
-0.054 943
7,3
4,3
1,392 408
0,527 153
0,166 756
0,050 988
0,026 514
0,001 975
-0,017 976
0,005 409
0,013 259
5,6
2,9
3,146 631
1,357 655
0,413 923
0,091 159
0,016 349
0,001 826
-0,00 4325
-0,00 4973
10,3
1,9
3.2.1. From 4,2 Kto the Triple Point ofNeon (24,5561 K)
with ""He as the Thermometric Gas. In this range Tgo is
defined by the relation:
a+ bp+ cp 2(4)
where p is the pressure in the gas thermometer and a, b
and c are coefficients the numerical values of which are
obtained from measurements made at the three defining
fixed points given in Sect. 3.2, but with the further restric-
tion that the lowest one of these points lies between 4,2 Kand 5,0 K.
3.2.2. From 3.0 Kto the Triple Point ofNeon (24,5561 K)with 3He or
AHe as the Thermometric Gas. For a 3He gas
thermometer, and for a4He gas thermometer used below
4,2 K, the non-ideality of the gas must be accounted for
explicitly, using the appropriate second virial coefficient
B3 {Tgo) or Bx (T90 ). In this range Tgo is defined by the
relation:
a + bp + cp 2
90 \+Bx(T90)N/V r }
where p is the pressure in the gas thermometer, a, b and
c are coefficients the numerical values of which are ob-
tained from measurements at three defining temperatures
as given in Sect. 3.2, N/V is the gas density with N being
the quantity of gas and V the volume of the bulb, x is 3
or 4 according to the isotope used, and the values of the
second virial coefficients are given by the relations:
For 3 He,
S(T90)/m3 mor' = {16,69-336.98 (T90/K)-
'
(6a)
+ 91.04(T9O /K)-2 -13.82(T9O/K)-
3] 1(T 6
.
For 4He.
B4 (T90)/m3 mo!
-1 = (16,708-374,05 (T90 /K)' (6b)
-383,53 (T90/K)-2 + 1799,2 (T90/K)-
3
-4033.2 (
r
90/K)-4 + 3252,8 (T90/K)-
5}10" b
Table 4. The constants A , At
\ B , Bx; C , C,; D and D-
tin the
reference functions of equations (9a); (9 b); (10a); and (10b) respec-
tively
AA,
A 2
A*A,A 6
A,
n 9
A lt
A l2
CoC,
C2
C 3
Qc
5
c6
c7
c8
C,
-2,135 347 29
3,183 247 20
-1,801 435 97
0,717 272 04
0,503 440 27
-0,618 993 95
-0,053 323 22
0,280 213 62
0,107 152 24
-0,293 028 65
0,044 598 72
0,118 686 32
-0,052 481 34
2,781 572 54
1,646 509 16
-0,137 143 90
-0,006 497 67
-0,002 344 44
0,005 118 68
0,001 879 82
-0,002 044 72
-0,000 461 22
0,000 457 24
B 0,183 324 722 B 13 -0,091 173 542
B, 0,240 975 303 B 14 0,001317 696
B 2 0,209 108 771 B 15 0,026 025 526
B3 0,190 439 972
B4 0,142 648 498
B5
0,077 993 465
B6 0,012 475 611
B7 -0,032 267 127
B8 -0,075 291 522
B9 -0,056 470 670
B 10 0,076 201 285
B,, 0,123 893 204
B 12 -0,029 201 193
D 439,932 854
D, 472,418 020
D, 37,684 494
D\ 7,472 018
04 2,920 828
D5
0,005 184
D 6 -0,963 864
D 1-0,188 732
0,191 203
0,049 025
The accuracy with which T90 can be realized using Eqs. (4)
and (5) depends on the design of the gas thermometer and
the gas density used. Design criteria and current goodpractice required to achieve a selected accuracy are given
in "Supplementary Information for the ITS-90".
3.3. The Triple Point of Equilibrium Hvdrogen
(13.8033 K) to the Freezing Point of Silver (961,78°C):
Platinum Resistance Thermometer
In this range T90 is defined by means of a platinum resis-
tance thermometer calibrated at specified sets of defining
fixed points, and using specified reference and deviation
functions for interpolation at intervening temperatures.
No single platinum resistance thermometer can
provide high accuracy, or is even likely to be usable, over
all of the temperature range 13,8033 K to 961,78 °C. Thechoice of temperature range, or ranges, from among those
listed below for which a particular thermometer can be
used is normally limited by its construction.
For practical details and current good practice, in
particular concerning types of thermometer available,
their acceptable operating ranges, probable accuracies,
permissible leakage resistance, resistance values, andthermal treatment, see "Supplementary Information for
the ITS-90". It is particularly important to take account
of the appropriate heat treatments that should be fol-
lowed each time a platinum resistance thermometer is
subjected to a temperature above about 420 °C.
Temperatures are determined in terms of the ratio of
the resistance R (T90 ) at a temperature T90 and the resis-
tance R (273,16 K) at the triple point of water. This ratio,
W(T90 ), is2
:
W(T90 ) = R(T90)/R (273,16 K)
.
(7)
An acceptable platinum resistance thermometer mustbe made from pure, strain-free platinum, and it must sat-
isfy at least one of the following two relations:
W (29,7646 °C)> 1,1 18 07,
If(-38,8344 oC)<0,844 235.
(8 a)
(8 b)
An acceptable platinum resistance thermometer that
is to be used up to the freezing point of silver must also
satisfy the relation:
W (961,78 °C)> 4,2844. (8 c)
In each of the resistance thermometer ranges, T90 is
obtained from Wr(T90 ) as given by the appropriate refer-
ence function {Eqs. (9b) or (10b)}, and the deviation
W(T90)— Wr(T90 ). At the defining fixed points this devia-
tion is obtained directly from the calibration of the ther-
mometer: at intermediate temperatures it is obtained by
means of the appropriate deviation function {Eqs. (12),
(13) and (14)}.
(i) - For the range 13,8033 K to 273,16 K the following
reference function is defined:
\n[WT(T90 )] = A Q +Y. A,
'ln(T90/273,16K) + l,5'
15(9 a)
An inverse function, equivalent to Eq. (9 a) to within
0,1 mK, is:
T9O/273,16K = B +I B-,
Wr(T90 )
1 16 -0,65
0.35(9 b)
The values of the constants A , A{, B and B, are given in
Table 4.
A thermometer may be calibrated for use throughout
this range or, using progressively fewer calibration points,
for ranges with low temperature limits of 24,5561 K,
54,3584 K and 83,8058 K, all having an upper limit of
273,16 K.
(ii) - For the range 0°C to 961,78 °C the following refer-
ence function is defined:
WAT90 ) = C +X C,T90/K- 754,1
5'
481(10a)
An inverse function, equivalent to equation (10 a) to with-
in 0,13 mK is:
T9O/K-273,15 = D +X; D,X(T90)-2,64
"
1,64(10b)
The values of the constants C , Cx, D and D, are given in
Table 4.
2 Note that this definition of W{Tqo ) differs from the corresponding
definition used in the ITS-27, ITS-48. IPTS-48, and IPTS-68: for all
of these earlier scales W{T) was defined in terms of a reference
temperature of 0°C, which since 1954 has itself been defined as
273,15 K
A thermometer may be calibrated for use throughout
this range or, using fewer calibration points, for ranges
with upper limits of 660,323°C, 419,527°C, 231,928°C,
1 56,5985 °C or 29,7646 °C, all having a lower limit of °C.
(iii) - A thermometer may be calibrated for use in the
range 234,3156 K (- 38,8344 °C) to 29,7646 °C, the cali-
bration being made at these temperatures and at the triple
point of water. Both reference functions {Eqs. (9) and
(10)} are required to cover this range.
The defining fixed points and deviation functions for
the various ranges are given below, and in summary form
in Table 5.
3.3.1. The Triple Point of Equilibrium Hydrogen
(13,8033 K) to the Triple Point of Water (273,16 K).
The thermometer is calibrated at the triple points of equi-
librium hydrogen (13,8033 K), neon (24,5561 K), oxygen
(54,3584 K), argon (83,8058 K), mercury (234,3156 K),
and water (273,16 K), and at two additional temperatures
close to 1 7,0 K and 20,3 K. These last two may be deter-
mined either: by using a gas thermometer as described in
Sect. 3.2, in which case the two temperatures must lie
within the ranges 16,9 K to 17,1 K and 20,2 K to 20,4 Krespectively; or by using the vapour pressure-temperature
relation of equilibrium hydrogen, in which case the twotemperatures must lie within the ranges 17,025 K to
17,045 K and 20,26 K to 20,28 K respectively, with the
precise values being determined from Eqs. (11a) and (lib)
respectively:
T90/K -1 7,035 =(p/kPa- 33,321 3)/l 3,32 (11a)
T90/K- 20,27 =(p/kPa-101,292)/30. (lib)
The deviation function is3
:
Table 5. Deviation functions and calibration points for platinum
resistance thermometers in the various ranges in which they define
W{T90)-W r(Tqo ) = a[W(T90)-\] + b[W(T90)-\]
2
+ 1 Ci [\nW(T90)r" (12)
with values for the coefficients a, b and csbeing obtained
from measurements at the defining fixed points and with
« = 2.
For this range and for the sub-ranges 3.3.1.1 to 3.3.1.3
the required values of Wr(T90 ) are obtained from Eq. (9a)
or from Table 1.
3.3.1.1. The Triple Point of Neon (24,5561 K) to the
Triple Point of Water (273,16 K). The thermometer is
calibrated at the triple points of equilibrium hydrogen(13,8033 K), neon (24,5561 K), oxygen (54,3584 K), argon
(83,8058 K), mercury (234,3156 K) and water (273,16 K).
The deviation function is given by Eq. (12) with values
for the coefficients a, b, cl , c 2
and c 3being obtained from
measurements at the defining fixed points and with
c4= c5= « = 0.
3.3.1.2. The Triple Point of Oxygen (54.3584 K) to the
Triple Point of Water (273,16 K). The thermometer is
a Ranges with an upper limit of 273,16 K
Sec- Lower Deviation functions
tion temper-
ature limit
T/K.
Calibration
points
(see Table 1)
3.3.1 13,8033 a[W(T90)-i] + b[W(T90)-l]2 2-9
+ £ Ci nnW(T90)r", n = 2i= i
3.3.1.1 24,5561 As for 3.3.1 with c4 = c 5= n = 2,5-9
3.3.1.2 54,3584 As for 3.3.1 with 6-9C2 =C3=C4. = C5=0, H=l
3.3.1.3 83,8058 a[W(T90)-\] 7-9
+ b[W(T90)-\]\nW(T90)
b Ranges with a lower limit of0°C
Sec- Upper Deviation functions Calibration
tion temper- points
ature limit (see Table 1
)
r/°C
3.3.2* 961,78 a[W(r90)-i]+fe[w(r90)-i]2
+ c[W(T9Q)-\f + d[W(T90)
-W (660,323 °C)]2
9,12-15
3.3.2.1 660,323 As for 3.3.2 with d=0 9,12-14
3.3.2.2 419,527 As for 3.3.2 with c = d = 9,12,13
3.3.2.3 231,928 As for 3.3.2 with c = d = 9,11,12
3.3.2.4 156,5985 As for 3.3.2 with b = c = d = Q 9,11
3.3.2.5 29,7646 As for 3.3.2 with fe = c = <f = 9,10
3 This deviation function {and also those of Eqs. (13) and (14)} maybe expressed in terms of W
rrather than W\ for this procedure see
"Supplementary Information for ITS-90"
c Range from 234,3156 K ( -38,8344°C) to 29. 7646°C
3.3.3 As for 3.3.2 with c= d=0 8-10
* Calibration points 9, 12-14 are used with d=0 for ;90 <;
660.323°C; the values of a. b and c thus obtained are retained for
/90 > 660.323 °C, with d being determined from calibration point 1
5
calibrated at the triple points of oxygen (54,3584 K),
argon (83,8058 K), mercury (234,3156 K) and water
(273,16 K).
The deviation function is given by Eq. (12) with values
for the coefficients a, b and clbeing obtained from mea-
surements at the defining fixed points, with c 2 =c 3=
c4 = c5= and with n = 1
.
3.3.1.3. The Triple Point of Argon (83.8058 K) to the
Triple Point of Water (273,16 K). The thermometer is
calibrated at the triple points of argon (83,8058 K), mer-
cury (234,3156 K) and water (273,16 K).
The deviation function is:
W(T90)- WT(T90 ) = a [W{T90)-\]
+ b[W(T90)-\] In W(T90 ) (13)
with the values of a and b being obtained from measure-
ments at the defining fixed points.
3.3.2. From 0°C to the Freezing Point of Silver
(961,78°C). The thermometer is calibrated at the triple
point of water (0,01 °C), and at the freezing points of tin
(231,928 °C), zinc (419,527 °C), aluminium (660,323 °C)
and silver (961,78 °C).
The deviation function is:
W(T90)-Wr(T90 ) = a[W(T90)-\] + b[W(T90)-\]
2(14)
+ c[W(T90)-\]° + d[W(T90)-W (660,323 °C)]2
.
For temperatures below the freezing point of aluminiumd= 0, with the values of a, b and c being determined from
the measured deviations from Wr(T9Q) at the freezing
points of tin, zinc and aluminium. From the freezing point
of aluminium to the freezing point of silver the above
values of a, b and c are retained and the value of d is
determined from the measured deviation from Wr(T90 ) at
the freezing point of silver.
For this range and for the sub-ranges 3.3.2.1 to 3.3.2.5
the required values for WT(Tgo ) are obtained from
Eq. (10a) or from Table 1.
3.3.2.1. From 0°C to the Freezing Point of Aluminium
(660.323° C). The thermometer is calibrated at the triple
point of water (0,01 °C), and at the freezing points of tin
(231,928 °C), zinc (41 9,527 °Q and aluminium (660,323 °C).
The deviation function is given by Eq. (14), with the
values of a, b and c being determined from measurements
at the defining fixed points and with d = 0.
3.3.2.2. From 0°C to the Freezing Point of Zinc
(419,527°C). The thermometer is calibrated at the triple
point of water (0,01 °C), and at the freezing points of tin
(231,928°C) and zinc (419,527°C).
The deviation function is given by Eq. (14) with the
values of a and b being obtained from measurements at
the defining fixed points and with c = d = 0.
3.3.2.3. From 0°C to the Freezing Point of Tin
(231.928°C). The thermometer is calibrated at the triple
point of water (0,01 °C), and at the freezing points of indi-
um (156.5985°C) and tin (231,928 °C).
The deviation function is given by Eq. (14) with the
values of a and b being obtained from measurements at
the defining fixed points and with c = </= 0.
3.3.2.4. From 0°C to the Freezing Point of Indium
(156.5985°C). The thermometer is calibrated at the
triple point of water (0.01 °C), and at the freezing point of
indium (1 56,5985 "C).
The deviation function is given by Eq. (14) with the
value of a being obtained from measurements at the defin-
ing fixed points and with b = c = d = 0.
3.3.2.5. From 0°C to the Melting Point of Gallium
(29.7646°C). The thermometer is calibrated at the triple
point of water (0,01 °C), and at the melting point of galli-
um (29,7646 °C).
The deviation function is given by Eq. (14) with the
value of a being obtained from measurements at the defin-
ing fixed points and with b = c = d = 0.
3.3.3. The Triple Point of Mercury (- 38.8344
J
C) to the
Melting Point of Gallium (29. 7646° C). The thermometer
is calibrated at the triple points of mercury ( — 38,8344 °C),
and water (0,01 °C), and at the melting point of gallium
(29,7646 °C).
The deviation function is given by Eq. (14) with the
values of a and b being obtained from measurements at
the defining fixed points and with c= d= 0.
The required values of WT(T90 ) are obtained from
Eqs. (9 a) and (10 a) for measurements below and above273,16 K respectively, or from Table 1.
3.4. The Range Above the Freezing Point of Silver
(961. 78°C): Planck Radiation Law
Above the freezing point of silver the temperature T90 is
defined by the equation:
L X (T9I(15)^
exptcJAT^tX)]- 1)-!
L x [T90 (X)] exp(c 2 [*T90]~l)-\
'
where T90 (X) refers to any one of the silver {
T
90 (Ag)= 1234,93 K}, the gold {
T
90 (Au)=l 337,33 K} or the
copper {T90 (Cu)= 1357,77 K} freezing points* and in
which L X {T90 ) and L X [T90 (X)] are the spectral concen-
trations of the radiance of a blackbody at the wavelength
(in vacuo) k at T90 and at T90 (X) respectively, andc2 =0,014388m-K.
For practical details and current good practice for
optical pyrometry, see "Supplementary Information for
the ITS-90" (BIPM-1990).
4. Supplementary Information and Differences
from Earlier Scales
The apparatus, methods and procedures that will serve to
realize the ITS-90 are given in "Supplementary Informa-
tion for the ITS-90". This document.also gives an account
of the earlier International Temperature Scales and the
numerical differences between successive scales that in-
clude, where practicable, mathematical functions for the
differences T90 — T68 . A number of useful approximations
to the ITS-90 are given in "Techniques for Approximating
the ITS-90".
These two documents have been prepared by the
Comite Consultatif de Thermometrie and are published
by the BIPM; they are revised and updated periodically.
The differences T90 — T68 are shown in Fig. 1 andTable 6. The number of significant figures given in Table 6
allows smooth interpolations to be made. However, the
reproducibility of the IPTS-68 is, in many areas, substan-
tially worse than is implied by this number.
4 The T90 values of the freezing points of silver, gold and copper are
believed to be self consistent to such a degree that the substitution
of any one of them in place of one of the other two as the reference
temperature Tqo (X) will not result in significant differences in the
measured values of Ton .
Table 6. Differences between ITS-90 and EPT-76, and between ITS-90 and IPTS-68 for specified values of T90 and t90
(T90-T76 )/mK
T90/K 1 2 3 4 5 6 7 8 9
-0,1 -0,2 -0,3 -0,4 -0,5
10 -0,6 -0,7 -0,8 -1,0 -1,1 -1,3 -1,4 -1,6 -1,8 -2,0
20 -2,2 -2,5 -2,7 -3,0 -3,2 -3,5 -3,8 -4,1
(T9Q --T6S)/K
r90/K 1 2 3 4 5 6 7 8 9
10 -0,006 -0,003 -0,004 -0,006 -0,008 -0,009
20 -0,009 -0,008 -0,007 -0,007 -0,006 -0,005 -0,004 -0,004 -0,005 -0,006
30 -0,006 -0,007 -0,008 -0,008 -0,008 -0,007 -0,007 -0,007 -0,006 -0,006
40 -0,006 -0,006 -0,006 -0,006 -0,006 -0,007 -0,007 -0,007 -0,006 -0,006
50 -0,006 -0,005 -0,005 -0,004 -0,003 -0,002 -0,001 0,000 0,001 0,002
60 0,003 0,003 0,004 0,004 0,005 0,005 0,006 0,006 0,007 0,007
70 0,007 0,007 0,007 0,007 0,007 0,008 0,008 0,008 0,008 0,008
80 0,008 0,008 0,008 0,008 0,008 0,008 0,008 0,008 0,008 0,008
90 0,008 0,008 0,008 0,008 0,008 0,008 0,008 0,009 0,009 0,009
T90/K 10 20 30 40 50 60 70 80 90
100 0,009 0,011 0,013 0,014 0,014 0,014 0,014 0,013 0,012 0,012
200 0,011 0,010 0,009 0,008 0,007 0,005 0,003 0,001
('90 -t68)/°c
t9o/°C -10 -20 -30 -40 -50 -60 -70 -80 -90
-100 0,013 0,013 0,014 0,014 0,014 0,013 0,012 0,010 0,008 0.008
0,000 0,002 0,004 0,006 0,008 0,009 0,010 0,011 0,012 0,012
'9o/°C- 10 20 30 40 50 60 70 80 90
0,000 -0,002 -0,005 -0,007 -0,010 -0,013 -0,016 -0,018 -0.021 -0,024
100 -0,026 -0,028 -0,030 -0,032 -0,034 -0,036 -0,037 -0,038 -0,039 -0,039
200 -0,040 -0,040 -0,040 -0,040 -0,040 -0,040 -0,040 -0,039 -0,039 -0,039
300 -0.039 -0,039 -0,039 -0,040 -0,040 -0,041 -0,042 -0,043 -0,045 -0,046
400 -0.048 -0,051 -0,053 -0,056 -0,059 -0,062 -0,065 -0,068 -0,072 -0,075
500 -0,079 -0,083 -0,087 -0,090 -0,094 -0,098 -0,101 -0,105 -0,108 -0,112
600 -0.115 -0,118 -0,122 -0,125* -0,08 -0,03 0,02 0,06 0,11 0,16
700 0,20 0,24 0,28 0.31 0,33 0,35 0,36 0,36 0,36 0,35
800 0,34 0,32 0,29 0.25 0,22 0,18 0,14 0,10 0,06 0,03
900 -0,01 -0,03 -0,06 -0,08 -0.10 -0,12 -0,14 -0,16 -0,17 -0,18
1000 -0,19 -0.20 -0,21 -0,22 -0.23 -0,24 -0,25 -0,25 -0,26 -0,26
ho/°C 100 200 300 400 500 600 700 800 900
1000 -0,26 -0.30 -0,35 -0,39 -0,44 -0,49 -0.54 -0,60 -0.66
2000 -0,72 -0,79 -0.85 -0.93 -1,00 -1,07 -1,15 -1.24 -1,32 -1,41
3000 -1.50 -1,59 -1.69 -1,78 -1.89 -1,99 -2,10 -2,21 -2,32 -2,43
A discontinuity in the first derivative of (r90 —
1
68 ) occurs at a temperature of t90 = 630,6°C,at which (t90 — f6s)= — 0,125eC
Appendix
The International Temperature Scale of 1927 (ITS-27)
The International Temperature Scale of 1927 was adopted by the
seventh General Conference of Weights and Measures to overcomethe practical difficulties of the direct realization of thermodynamictemperatures by gas thermometry, and as a universally acceptable
replacement for the differing existing national temperature scales.
The ITS-27 was formulated so as to allow measurements of temper-
ature to be made precisely and reproducibly, with as close an ap-
proximation to thermodynamic temperatures as could be deter-
mined at that time. Between the oxygen boiling point and the gold
freezing point it was based upon a number of reproducible temper-
atures, or fixed points, to which numerical values were assigned, andtwo standard interpolating instruments. Each of these interpolating
instruments was calibrated at several of the fixed points, this giving
the constants for the interpolating formula in the appropriate tem-
perature range. A platinum resistance thermometer was used for the
lower part and a platinum rhodium/platinum thermocouple for
temperatures above 660 °C. For the region above the gold freezing
point, temperatures were defined in terms of the Wien radiation law:
in practice, this invariably resulted in the selection of an optical
pyrometer as the realizing instrument.
The International Temperature Scale of 1948 (ITS-48)
The International Temperature Scale of 1948 was adopted by the
ninth General Conference. Changes from the ITS-27 were: the low-
er limit of the platinum resistance thermometer range was changed
from — 190°C to the defined oxygen boiling point of — 182.97°C.
and the junction of the platinum resistance thermometer range and
the thermocouple range became the measured antimony freezing
point (about 630°C) in place of 660"C; the silver freezing point
was defined as being 960.8 C instead of 960.5'C; the gold freezing
point replaced the gold melting point (1063"C); the Planck radia-
tion law replaced the Wien law: the value assigned to the second
radiation constant became 1.438 x I0~ : m K. in place of 1.432 x
10"" m • K; the permitted ranges for the constants of the interpola-
10
tion formulae for the standard resistance thermometer and thermo-
couple were modified; the limitation on XT for optical pyrometry
(A7"<3 x 10~ 3 m K.) was changed to the requirement that "visible"
radiation be used.
The International Practical Temperature Scale
of 1948 (Amended Edition of 1960) (IPTS-48)
The International Practical Temperature Scale of 1948, amended
edition of 1960, was adopted by the eleventh General Conference:
the tenth General Conference had already adopted the triple point
of water as the sole point defining the kelvin, the unit of thermody-
namic temperature. In addition to the introduction of the word
"Practical", the modifications to the ITS-48 were: the triple point
of water, defined as being 0,01 °C, replaced the melting point of ice
as the calibration point in this region; the freezing point of zinc,
defined as being 419,505 °C, became a preferred alternative to the
sulphur boiling point (444,6 °C) as a calibration point; the permitted
ranges for the constants of the interpolation formulae for the stan-
dard resistance thermometer and the thermocouple were further
modified; the restriction to "visible" radiation for optical pyrome-
try was removed.
Inasmuch as the numerical values of temperature on the IPTS-
48 were the same as on the ITS-48. the former was not a revision of
the scale of 1948 but merely an amended form of it.
The International Practical Temperature Scale
of 1968 (IPTS-68)
In 1968 the International Committee of Weights and Measures
promulgated the International Practical Temperature Scale of 1968,
having been empowered to do so by the thirteenth General Confer-
ence of 1967-1968. The IPTS-68 incoporated very extensive
changes from the IPTS-48. These included numerical changes, de-
signed to bring it more nearly in accord with thermodynamic tem-
peratures, that were sufficiently large to be apparent to many users.
Other changes were as follows: the lower limit of the scale was
extended down to 13,81 K: at even lower temperatures (0,5 K to
5.2 K), the use of two helium vapour pressure scales was recom-
mended: six new defining fixed points were introduced - the triple
point of equilibrium hydrogen (13,81 K), an intermediate equilibri-
um hydrogen point (17,042 K.). the normal boiling point of equilib-
rium hydrogen (20,28 K), the boiling point of neon (27.102 K), the
triple point of oxygen (54,361 K), and the freezing point of tin
(231.9681 °C) which became a permitted alternative to the boiling
point of water; the boiling point of sulphur was deleted; the values
assigned to four fixed points were changed - the boiling point of
oxygen (90.188 K), the freezing point of zinc (419,58°C), the freez-
ing point of silver (961,93 °C), and the freezing point of gold
( 1064,43 °C); the interpolating formulae for the resistance ther-
mometer range became much more complex; the value assigned to
the second radiation constant c 2 became 1,4388 x 10~ 2 m • K; the
permitted ranges of the constants for the interpolation formulae for
the resistance thermometer and thermocouple were again modified.
The International Practical Temperature Scale
of 1968 (Amended Edition of 1975) (IPTS-68)
The International Practical Temperature Scale of 1968, amendededition of 1975, was adopted by the fifteenth General Conference in
1975. As was the case for the IPTS-48 with respect to the ITS-48,
the IPTS-68 (75) introduced no numerical changes. Most of the
extensive textural changes were intended only to clarify and simplify
its use. More substantive changes were: the oxygen point was de-
fined as the condensation point rather than the boiling point; the
triple point of argon (83,798 K) was introduced as a permitted
alternative to the condensation point of oxygen; new values of the
isotopic composition of naturally occurring neon were adopted; the
recommendation to use values of Tgiven by the 19584He and 1962
3He vapour-pressure scales was rescinded.
The 1976 Provisional 0,5 K to 30 K Temperature Scale
(EPT-76)
The 1976 Provisional 0.5 K. to 30 K. Temperature Scale was intro-
duced to meet two important requirements: these were to provide
means of substantially reducing the errors (with respect to corre-
sponding thermodynamic values) below 27 K that were then knownto exist in the IPTS-68 and throughout the temperature ranges of
theaHe and 3He vapour pressure scales of 1958 and 1962 respective-
ly, and to bridge the gap between 5,2 K and 13,81 K in which there
had not previously been an international scale. Other objectives in
devising the ETP-76 were "that it should be thermodynamically
smooth, that it should be continuous with the IPTS-68 at 27,1 K.
and that is should agree with thermodynamic temperature T as
closely as these two conditions allow". In contrast with the IPTS-68.
and to ensure its ranid adoption, several methods of realizing the
ETP-76 were appr ved. These included: using a thermodynamicinterpolation instrument and one or more of eleven assigned refer-
ence points: taking differences from the IPTS-68 above 13.81 K;
taking differences from helium vapour pressure scales below 5 K:
and taking differences from certain well-established laboratory
scales. Because there was a certain "lack of internal consistency" it
was admitted that "slight ambiguities between realizations" might
be introduced. However the advantages gained by adopting the
EPT-76 as a working scale until such time as the IPTS-68 should be
revised and extended were considered to outweigh the disadvan-
tages.
6.1.2 Reproduction of the article by T. J. Quinn, entitled "News from the BIPM"
,
published in Me tro 1ogia 26, 69-74 (1989).
99
Metrologia 26, 69-74 (1989) metrologia© Springer-Verlag 1989
News from the BIPMT.J. Quinn
Bureau International des Poids et Mesures, Pavilion de Breteuil, F-92312 Sevres Cedex, France
Received: October 14, 1988
Comite International des Poids et Mesures
77th Meeting
The Comite International des Poids et Mesures(CIPM) at its 77th Meeting, held at the Pavilion de
Breteuil on the 4th, 5th and 6th of October 1988,
adopted two important Recommendations concern-
ing the use of the Josephson effect and the quantumHall effect for maintaining reference standards for the
measurement of emf and resistance. These Recommen-dations are based upon proposals made to the CIPMby its Comite Consultatif d' Electricite (CCE) which
met in September 1988. The CCE proposals were the
result of a great deal of work and discussion that hadtaken place among representatives of the national
standards laboratories and the BIPM, particularly
over the past twelve months.
New determinations of the SI volt and ohm, direct-
ly by realizations of the SI definitions and indirectly
through determinations of the relevant fundamental
physical constants, have established the values of the
Josephson constant K, and the von Klitzing constant
RK of the quantum Hall effect with uncertainties of a
few parts in 107
. The constants K, and RK were as-
sumed by the CCE to be equal to 2 e/h and h/e2respec-
tively for the purpose of including determinations of
fundamental constants in their evaluation. The repro-
ducibility, however, of the emf across a Josephson
junction and the resistance of a quantum Hall device
is much better than a few parts in 10 7 and approaches
one part in 108
. Furthermore, the value of 2 e/h sug-
gested by the CCE in 1972 is now known to be in error
by about 8 parts in 106 , although not all national
standards laboratories are using this value.
In order to unify world reference standards of emfand allow users to take advantage of the great repro-
ducibility of the Josephson effect and the quantumHall effect, the CIPM acted on the advice of the CCEand adopted particular values of K, and R K , desig-
nated Kj_ qo and R K -Q < f° r use by all laboratories
beginning on 1st January 1990. For those countries
Reprinted with, permission of BIPM.
that have based their reference standard of emf on the
Josephson junction and use the 1972 CCE value for
2 e/h, the adoption of the new value for K, will lead to
a change in their reference standards of about 8 parts
in 10 6, i. e. 8 uV per volt. For the countries using values
of K, which differ from the 1972 CCE value, the
changes will be within the range of about 9 to 3.5 parts
in 10 6.
Also 1st January 1990 it is expected that a newinternational temperature scale, the International
Temperature Scale of 1990 (ITS-90), will come into
effect, which will replace the International Practical
Temperature Scale of 1968 (IPTS-68). The differences
between ITS-90 and IPTS-68 are significant; not only
are the scales substantially different in definition but
temperatures measured on the ITS-90 differ signifi-
cantly from those measured on the IPTS-68. For ex-
ample, near room temperature: for t68 ^ 20 °C the dif-
ference T90 - T68 ~ — 5 mK and for t68 ~ 100 °C the
difference T90 — T68 ^ — 25 mK. The normal boiling
point of water thus will be about 99.97 °C and not
100.00 °C as in the past.
It is important that users be made aware of these
changes in electrical and temperature reference stan-
dards well in advance of their introduction. For this
reason the 18th Conference Generale des Poids et
Measures asked the CIPM to announce one year in
advance, namely at the beginning of 1989, the magni-tude of the changes consequent upon their coming into
effect.
These official announcements are now made in the
following three CIPM recommendations:
Representation of the volt by meansof the Josephson effect
Recommendation 1 (CI-1988)
The Comite International des Poids et Mesures,
acting in accordance with instructions given in Reso-
lution 6 of the 18th Conference Generale des Poids et
Mesures concerning the forthcoming adjustment of
the representations of the volt and the ohm,
considering
-that a detailed study of the results of the
most recent determinations leads to a value of
483 597.9 GHz/V for the Josephson constant, K,, that
is to say, for the quotient of frequency divided by the
potential difference corresponding to the n = 1 step in
the Josephson effect,
- that the Josephson effect together with this value
of ATj can be used to establish a reference standard of
electromotive force having a one-standard-deviation
uncertainty with respect to the volt estimated to be
4 parts in 107
, and a reproducibility which is signifi-
cantly better,
recommends- that 483 597.9 GHz/V exactly be adopted as a
conventional value, denoted by K^ 90 , for the Joseph-
son constant, Klt
-that this new value be used from 1st
January 1990, and not before, to replace the values
currently in use,
- that this new value be used from this same date
by all laboratories which base their measurements of
electromotive force on the Josephson effect, and- that from this same date all other laboratories
adjust the value of their laboratory reference stan-
dards to agree with the new adopted value,
is of the opinion
- that no change in this recommended value of the
Josephson constant will be necessary in the foresee-
able future, and
draws the attention of laboratories to the fact that the
new value is greater by 3.9 GHz/V, or about 8 parts in
106 , than the value given in 1972 by the Comite Consul-
tatif d 'Electricite in its Declaration E-72.
Representation of the ohm by means
of the quantum Hall effect
Recommendation 2 (CI- (988)
The Comite International des Poids et Mesures,
acting in accordance with instructions given in
Resolution 6 of the 18th Conference Generale des
Poids et Mesures concerning the forthcoming adjust-
ment of the representations of the volt and the ohm,
considering
- that most existing laboratory reference standards
of resistance change significantly with time.
- that a laboratory reference standard of resistance
based on the quantum Hall effect would be stable and
reproducible,
- that a detailed study of the results of the most
recent determinations leads to a value of 25 812.807 Q.
for the von Klitzing constant, RK , that is to say, for
the quotient of the Hall potential difference divided
by current corresponding to the plateau i = 1 in the
quantum Hall effect,
- that the quantum Hall effect, together with this
value of RK , can be used to establish a reference stan-
dard of resistance having a one-standard-deviation
uncertainty with respect to the ohm estimated to be
2 parts in 107
, and a reproducibility which is signifi-
cantly better,
recommends- that 25 812.807 Q. exactly be adopted as a con-
ventional value, denoted by RK-9o> f°r tne von Klit-
zing constant, RK ,
-that this value be used from 1st January 1990,
and not before, by all laboratories which base their
measurements of resistance on the quantum Hall effect,
- that from this same date all other laboratories
adjust the value of their laboratory reference stan-
dards to agree with RK-90 ,
- that in the use of the quantum Hall effect to
establish a laboratory reference standard of resistance,
laboratories follow the most recent edition of the
technical guidelines for reliable measurements of the
quantized Hall resistance drawn up by the Comite
Consultatif d 'Electricite and published by the Bureau
International des Poids et Mesures, and
is of the opinion
- that no change in this recommended value of the
von Klitzing constant will be necessary in the foresee-
able future.
Preparation of the International Temperature Scale
of 1990 (ITS-90)
Recommendation 3 (CI-1988)
The Comite International des Poids et Mesures,
acting in accordance with instructions given in
Resolution 7 of the 18th Conference Generale des
Poids et Mesures concerning the preparation of the
new international temperature scale,
announces that the differences between the ITS-90 and
the IPTS-68 will be approximately those indicated in
the graph attached to this Recommendation,
recommends that national laboratories take note of
these differences with a view to the implementation of
the ITS-90 on 1st January 1990.
0.02-
J> -0.02 -
*? -0.04
23
2<5
Q.
E -0.2
-i 1 1 1 1 r
- 0.4
0.2
J_ J_ _L
0.2
-200
Fig. 1. Graph of r 90- t68
200 400
f/°C
600 800 1000
Comite Consultatif d'Electricite
18th Meeting
The CCE at its meeting in September 1988 consid-
ered very carefully the way in which the recommendedvalues for K, and RK (defined above) be used. Al-
though no formal recommendation was made on this
point, the following four statements were drafted and
unanimously approved by the CCE and subsequently
supported by the CIPM '
:
/./. Recommendations 1 (CI-1988) and 2 (CI-1988)
do not constitute a redefinition of SI units
The conventional values K,- go and RK - go cannot be
used as bases for defining the volt and ohm (meaning
the present volt and ohm units in the Systeme Inter-
national d'Unites). To do so would change the status
ofn from that of a constant having an exactly defined
value (and would therefore abrogate the definition of
the ampere) and would also produce electrical units
which would be incompatible with the definition of the
kilogram and units derived from it.
Because the original draft CCE recommendations were slighly
modified by the CIPM, I have chosen to refer here only to the
CIPM designations, i.e., CI-1988.
1.2. Concerning the use of subscripts on symbols
for quantities or units
The CCE considers that symbols for electromotive
force (electric potential, electric potential difference)
and electric resistance, and for the volt and the ohm,
should not be modified by adding suffixes to denote
particular laboratories or dates.
The principal reasons for this viewpoint with re-
spect to the physical quantities are that:
- until now, temperature being one of the very few
exceptions, it has not been necessary to introduce ex-
plicitly the concept of a system of conventional phys-
ical quantities differing from the traditional quanti-
ties,
-it would be difficult to make such a concept
widely understood and accepted,
- the concept, if introduced for electromotive force
and electrical resistance, would propagate to other
quantities.
The principal reasons for this viewpoint with re-
spect to the units are that:
- the appearance of creating a unit system other
than SI should be avoided, particularly as this would
propagate to units for other quantities.
- the new reference standards will be completely
satisfactory representations of the volt and the ohmfor the great majority of applications,
- any disagreement between those laboratories
that realize the new reference standards will be negli-
gible from the point of view of the great majority of
users,
- many countries are in any case constrained by
their existing legislation concerning physical quanti-
ties and units to use the SI names and symbols.
1.3. Concerning the practical implementation of the
Recommendations 1 (CI-1988) and 2 (CI-1988)
The CCE having carefully considered the three ap-
proaches listed in the reports of the Working Groups,
documents CCE/88-34 and CCE/88-35 2, is of the
opinion that a rigorous solution to this problem has
been identified which avoids
(i) defining new units "V90" or "ft90'\
(ii) defining new physical quantities "£90" or "i?90"
.
(iii) the use of subscripts or other distinguishing sym-
bols of any sort on either unit symbols or quantity
symbols.
The preferred approach is indicated in the following
example of a statement that may be communicated to
users of standard-cell calibration certificates
:
The measured emf, £, or electric potential differ-
ence, U, of the unknown source may be rigorously
expressed in terms of the SI volt, V, as:
£ = (1.018 xxx xx)V±e
The symbol e represents the total uncertainty, at the
level of one standard deviation, and is given by
£ = [(A£) 2 +(£- S)2 ]" 2
where A£ is the combined uncertainty in volts (at one
standard deviation) associated with the calibration it-
self and with the realization of the Josephson-effect
reference standard at the particular national standards
laboratory, and 5 represents the relative uncertainty
with which the ratio /Cj_ 90//C, is known. At present 8
is estimated to be 4 x 10" 7(one standard deviation)
according to Recommendation 1 (CI-1988).
Since, by international agreement, 5 is common to
all laboratories, it may be indicated separately and the
above expression for £ may be rewritten
£ = (1.018 xxx xx) V±A£
for all practical purposes of precision electrical metrol-
ogy and trade. However, when this is done, 8 should
Note to reader: these two Documents dealt with the Joseph-
son effect and the quantum Hail effect, respectively.
appear separately on the certificate where the preci-
sion is such as to require it. When A£/£ is significantly
greater than 4 x 10" 7, 8 may be omitted.
The treatment of resistance measurements [Rec-
ommendation 2 (CI-1988)] is strictly analogous.
1.4. Example of the wording to be used
on calibration certificates
The values of emf below are based on ... [a descrip-
tion of the calibration procedure may be placed
here] ... using the new conventional value of the
Josephson constant internationally adopted for use
from 1st January 1990 (see Note A).
cell number emf in volts uncertainty in volts
1 1.018 123 4 A£[other data related to the calibration may be placed
here]
Note AThe value of the Josephson constant used in this cali-
bration is K,_ 90 = 483 597.9 GHz/V and is that
adopted, by international agreement, for implementa-
tion on 1st January 1990 by all national standards
laboratories that use the Josephson effect as a refer-
ence standard of the volt. By international agreement,
all such laboratories now use the same value of the
Josephson constant whereas until recently they did
not. National standards laboratories of those coun-
tries that do not use the Josephson effect as a reference
standard can maintain their own reference standards
so as to be consistent with the above value of the
Josephson constant, by periodic comparisons with a
laboratory that does use the Josephson effect. An ideal
reference standard of emf based on the Josephson ef-
fect and Kj_ 90 is consistent with the SI volt within an
assigned fractional one-standard-deviation uncer-
tainty of 4 x 10" 7(0.4 uV for an emf of one volt).
Because this uncertainty is the same for all national
standards laboratories, it has not oeen formally in-
cluded in the uncertainties given in the table. However,
its existence must be recognized when the utmost
consistency between electrical and mechanical mea-
surement is required.
Cornice Consultatif pour la Masse
et les grandeurs apparencies
3rd Meeting
The Comite Consultatif pour la Masse et les gran-
deurs apparentees (CCM) met at the Pavilion de Bre-
teuil on the 26th and 27th of May 1988. The President,
Prof. Bray, being indisposed the meeting was chaired
by Dr. Giacomo, Director of the BIPM. The CCMcovers a wide range of activities in the fields of mass.
force and pressure standards and much of the meeting
was devoted to the examination of reports drawn upby the nine Working Groups. Most of these WorkingGroups had met in the days immediately preceding the
meeting of the CCM. The fields of work covered by
the Working Groups were: 1, the measurement of air
density; 2, the conservation and behaviour of Pt/Ir
mass standards; 3, the conservation and behaviour of
stainless-steel mass standards; 4, the measurement of
the density of liquids and solids; 5, force measurement;
6, 7, 8 and 9, the measurement of high, medium, low
and very low pressures, respectively.
In many of these fields, in addition to the individu-
al and collaborative projects under way, international
comparisons had taken place or were in progress and
reports were presented. The CCM also discussed
recent advances in high-accuracy weighing and bal-
ance design and the third verification of national
prototypes of the kilogram now in its preliminary
stages at the BIPM. It is expected that the first part of
this operation, the comparison of the international
prototype with its copies at the BIPM, will have been
completed by the end of 1988. Over the next three
years or so, national prototypes will come to the
BIPM in groups for verification. Since the last time
such a large scale verification took place, some forty
years ago, the accuracy of the best balances has im-
proved by at least a factor of ten. The present verifi-
cation is being carried out using the NBS-2 balance
which is a single-pan, knife-edge balance that has an
accuracy of about 1 ug in the comparison of Pt/Ir mass
standards, i.e. a relative uncertainty at the level of one
standard deviation of about 1 part in 109 .
Comite Consultatif pour les Etalons de Mesure
des Rayonnements Ionisants
Section I (X-rays, y-rays and electrons) and Sec-
tion III (Neutrons) of the Comite Consultatif pour
les Etalons de Mesure de Rayonnements Ionisants
(CCEMRI) met from 8th to 11th and 18th to 20th of
April respectively. During the meeting of Section I, the
work of the BIPM was reviewed and the results of
recent comparisons of national standards were dis-
cussed. The need for comparisons of absorbed-dose
standards for radiation of energy above 1 MeV wasstressed. Two Recommendations were adopted andsent to the CCEMRI for approval at its next meeting.
At the meeting of Section III, the BIPM work onneutron measurements was reviewed. The published
results of completed neutron-emission-rate and flu-
ence-rate comparisons, the NPL neutron-dosimetry
comparison and the preliminary report on the BIPMneutron-dosimetry comparison were all discussed.
Work in the BIPM Laboratories
Among the principal activities in the laboratories
of the BIPM over the past year have been the follow-
ing:
- Electricity section : a limited comparison of 1 Qnational reference standards of resistance was carried
out to check the consistency among the various real-
izations of the quantum Hall effect now being made in
1 1 national laboratories and the BIPM. Expressed in
terms of the BIPM's representation of the ohm,J269_ BI , on the central date of the comparison, the
measured results for the quantized Hall resistance in
five of the six laboratories claiming uncertainties of
4 parts in 108 or less are within a range of 7 parts in
10 8. The overall spread of all results was 6 parts in 10
7.
Refinements in the BIPM quantum-Hall-effect system,
based upon a cryogenic current comparator, now al-
low Q69 -si to be monitored with an uncertainty of 1.5
parts in 108.
Using Josephson junctions made from the newhigh-Tc superconducting material a measurement of
2e/h was made having an uncertainty of 3 parts in 106
and which differed from the value obtained from a
metallic Josephson junction by about 5 parts in 106
.
- Mass section: as already mentioned, work began
on the third international verification of national pro-
totypes of the kilogram. The BIPM flexure-strip bal-
ance has been used to carry out an experiment to
search for a nuclear-isospin-coupled "fifth-force". Noevidence for the existence of such a force was found,
but during the experiment weighings of 2.3 kg masses
were carried out that showed remarkably small stan-
dard deviations of about 5 parts in 1012
.
-Time section: as part of the regular analysis of
clock comparisons made using the GPS satellites
which go to make up International Atomic Time (TAI)
and Universal Coordinated Time (UTC), it was found
that for laboratories within about 1000 km of each
other errors in coordinates could be discerned and
corrected for with an accuracy of about 1 5 cm in the
geocentric coordinates x, y and z. As a result, time
comparisons between these (European) laboratories
can now be made to within a few nanoseconds.
- Laser section: important international compari-
sons of laser wavelengths and frequencies have been
carried out between a total of 8 countries, mostly at
wavelengths in the visible (A = 633 mm) but also one in
the infrared (X = 3.39 urn). Uncertainties in these com-
parisons are at level of about 1 part in 1011 and differ-
ences between lasers from participating countries were
found, in many cases, to be hardly greater than this.
- Ionizing radiations section : participation in the
international comparison of the measurements of ac-
tivity of iodine- 125 has begun. Two methods of spec-
trum analysis are being used in order to understand
better the results obtained earlier during a limited
comparison of measurement of activity of the same
nuclide. The long-term work on counting methods
and counting statistics continues with much emphasis
now being placed on the method of 'generalized dead
times'.
- Radiometry: new work has begun in radiometry.
Equipment has been built and is now in operation for
the comparison of calibrations of silicon diodes and
thermopiles. The method of self-calibration of silicon
diodes has been used and experience is being gained so
that international comparisons and, in due course,
calibrations can be carried out at the BIPM.
Other News
During the meeting of the CIPM in October 1988
a new building (of some 900 m2in surface) was inaugu-
rated. It will house a library and offices for scientific
staff, the secretariat and the Director. This completes
the second stage of a long-term plan for buildings at
the BIPM, the first stage of which was a new laborato-
ry building for laser work opened in 1984.
On 31st July 1988 Dr. P. Giacomo, who had been
Director of the BIPM since 1978, retired and was suc-
ceeded by Dr. T.J. Quinn, previously Deputy Direc-
tor.
Publications
Since October 1987 the BIPM has published:
18e Conference Generate des Poids et Mesures
(1987), Comptes Rendus, 108 pages.
Proces-verbaux des seances du Comite International
des Poids et Mesures, tome 55 (76esession, octobre
1987), 178 pages.
Comite Consultatif de Photometrie et Radiometrie,
llesession (1986), 183 pages.
Comite Consultatif de Thermometrie , 16esession
(1987), 162 pages.
Rapport Annuel du BIHpour 1987 (about 72 pages
covering the contribution of the BIPM on time scales).
Circulaire D du BIH (monthly) (contribution of
the BIPM on time scales).
Circulaire T (monthly), circular of the BIPMtaking over the part "Temps" of the circular D of the
BIH (Circular Tl for January 1988 was published on
March 1st, 1988).
Le BIPM et la Convention du Metre, illustrated
brochure in French and in English, 47 pages, widely
circulated through various national laboratories.
One should also mention some 40 articles pub-
lished by staff members in scientific journals and a
dozen BIPM reports; the complete list of these publi-
cations may be found in the Proces-Verbaux du
CIPM, 77th meeting (1 988), which will appear in 1 989.
Copies of these documents may be obtained upon
application to: Mr le Directeur du BIPM, Pavilion de
Breteuil, F-92312 Sevres Cedex, or from the Librairie
Offilib, 48 rue Gay-Lussac, F-75240 Paris Cedex 05.
6.1.3 Three recommendations were adopted by the CCT at its 17th Session [14].
These recommendations were considered by the CIPM and Recommendation Tl (1989)
of the CCT was adopted as Recommendation 5 (CI -89) of the CIPM. RecommendationsT2 (1989) and T3 (1989) of the CCT were noted by the CIPM as CCT recommendations.The CCT recommendations were as follows:
6.1.3.1 Recommendation Tl (1989)
The International Temperature Scale of 1990
The Comite Consultatif de Thermometrie (CCT) acting in accordance withResolution 7 of the 18 e CGPM has generated the International Temperature Scaleof 1990 (ITS -90) in order to supersede the International Practical TemperatureScale of 1968 (IPTS-68).
The CCT notes that, by comparison with the IPTS-68, the ITS -90- extends to lower temperatures, down to 0.65 K, and hence also supersedes
the EPT-76,is in substantially better agreement with corresponding thermodynamic
temperatures
,
has much improved continuity, precision and reproducibility throughoutits range and
- has subranges and alternative definitions in certain ranges which greatlyfacilitate its use.
The CCT also notes that, to accompany the text of the ITS -90 there will betwo further documents, the Supplementary Information for the ITS -90 andTechniques for Approximating the ITS -90. These documents will be published bythe BIPM and periodically updated.
The CCT recommendsthat on 1 January 1990 the ITS -90 come into force and
- that from this same date the IPTS-68 and the EPT-76 be abrogated.
106
6.1.3.2 Recommendation T2 (1989)
Reference tables for thermocouples and industrial platinum resistancethermometers
The Comite Consultatif de Thermometrie,considering
that the introduction of the International Temperature Scale of 1990(ITS-90) will lead to an urgent requirement for new reference tables for boththermocouples and industrial platinum resistance thermometers,
requests its Working Group 2
to collaborate with national laboratories in the rapid preparation ofnew reference tables taking into account not only the change from IPTS-68 to
ITS-90 but also new information on the behavior of thermocouples and industrialplatinum resistance thermometers,
recommendsthat these new tables be used as the basis for new national and
international reference tables for thermocouples and industrial platinumresistance thermometers and
that meanwhile the existing reference tables based upon IPTS-68 should beused in conjunction with the table of differences r90 - T68 which appears in the
ITS-90.
6.1.3.3 Recommendation T3 (1989)
The uncertainty inherent in the realization of the InternationalTemperature Scale of 1990
The Comite Consultatif de Thermometrie,considering the requirement for assigning an uncertainty to the numerical
value of any temperature on the International Temperature Scale of 1990 (ITS-90),encourages national laboratories to
a) quantify the uncertainties in the fixed point realizations,b) quantify the uncertainties resulting from the use of the specified
interpolating instruments of ITS-90,c) develop the mathematical procedures describing the propagation of these
uncertainties to any intermediate temperature
.
107
6.1.4 Reproduction of article by B. W. Mangum, entitled "Special Report on theInternational Temperature Scale of 1990; Report on the 17th Session of theConsultative Committee on Thermometry," and published in J. Res. Natl. Inst.Stand. Technol. 95, 69-77 (1990).
108
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
[J. Res. Natl. Inst. Stand. Technol. 95, 69 (1990)]
Special Report on the International Temperature Scale of1990
Report on the 1 7th Session of the
Consultative Committee on Thermometry
Volume 95 Number 1 January-February 1990
B. W. Mangum
National Institute of Standards
and Technology,
Gaithersburg, MD 20899
This article summarizes the results of
the 17th Session of the Consultative
Committee on Thermometry of the In-
ternational Committee of Weights andMeasures (Comite Consultatif de Ther-
mometrie of the Comite International
des Poids et Mesures) that met in
Sevres, France, September 12-14, 1989.
That session was devoted exclusively to
the completion of the International
Temperature Scale of 1990, described
herein, and to the implications of its
adoption.
Key words: CCT; Comite Consultatif de
Thermometrie; Consultative Committeeon Thermometry; International Tem-perature Scale of 1990; temperature;
temperature scale; thermodynamic tem-
perature; thermometry.
Accepted: November 4, 1989
1. Introduction
The Consultative Committee on Thermometry(Comite Consultatif de Thermometrie, CCT) is one
of eight Consultative Committees (Comites Con-
sultants) of the International Committee of
Weights and Measures (Comite International des
Poids et Mesures, CIPM). The CIPM is a commit-
tee of the General Conference of Weights and
Measures (Conference Generale des Poids et
Mesures, CGPM). The eight Consultative Commit-tees (Comites Consultatifs) of the CIPM are:
1. The Comite Consultatif d'Electricite (CCE),
established in 1927,
2. The Comite Consultatif de Photometrie et
Radiometric (CCPR), assigned this name in 1971;
the previous name was the Comite Consultatif de
Photometrie, established in 1933,
3. The Comite Consultatif de Thermometrie
(CCT), established in 1937,
4. The Comite Consultatif pour la Definition duMetre (CCDM), established in 1952,
5. The Comite Consultatif pour la Definition de
la Seconde (CCDS), established in 1956,
6. The Comite Consultatif pour les Etalons de
Mesure des Rayonnements Ionisants (CCEMRI),established in 1958,
7. The Comite Consultatif des Unites (CCU), es-
tablished in 1964, and
8. The Comite Consultatif pour la Masse et les
grandeurs apparentees (CCM), established in 1980.
The CCT is composed presently of members
from the following laboratories:
1. Amt fur Standardisierung, Messwesen und
Warenprufung [ASMW], Berlin, DDR,2. Bureau National de Metrologie, Paris,
France: Institut National de Metrologie, [INM] du
Conservatoire National des Arts et Metiers,
3. Ceskoslovensky Metrologicky Ustav [CSMU],
Bratislava, Czechoslovakia,
4. National Research Council [NRC], Ottawa,
Canada,
69
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
5. CSIRO, Division of Applied Physics
[CSIRO], Lindfield, Australia,
6. D. I. Mendeleyev Institute for Metrology
[VNIIM], Leningrad, USSR; Physico-Technical
and Radio-Technical Measurements Institute
(PRMI), Moscow, USSR,7. National Institute of Metrology [NIM], Bei-
jing, PRC,8. Istituto di Metrologia G. Colonnetti [IMGC],
Turin, Italy
9. Kamerlingh Onnes Laboratorium [KOL],
Leiden, The Netherlands,
10. Korea Standards Research Institute [KSRI],
Seoul, Korea,
11. National Institute of Standards and Technol-
ogy [NIST], Gaithersburg, MD, USA12. National Physical Laboratory [NPL], Ted-
dington, UK,13. National Research Laboratory of Metrology
[NRLM], Ibaraki, Japan,
14. Physikalisch-Technische Bundesanstalt [PTB],
Braunschweig, FRG,15. Van Swinden Laboratorium [VSL], Delft,
The Netherlands,
16. Iowa State University, Ames, Iowa, USA,and
17. Bureau International des Poids et Mesures
[BIPM], Sevres, France.
The CCT met September 12-14, 1989 at the Bu-
reau International des Poids et Mesures (BIPM) in
its 17th Session [1] and completed the final details
of the new temperature scale, the International
Temperature Scale of 1990 (ITS-90) [2]. The CCTthen recommended to the CIPM at its meeting on
September 26-28, 1989 at the BIPM that the ITS-
90 be adopted and made the official scale. TheCIPM did adopt the new temperature scale at their
meeting [3] and the ITS-90 became the official in-
ternational temperature scale on January 1, 1990,
the same date on which changes affecting certain
electrical reference standards were implemented
[4]. The ITS-90 supersedes the International Practi-
cal Temperature Scale of 1968, Amended Edition
of 1975 [IPTS-68(75)] [5] and the 1976 Provisional
0.5 K to 30 K Temperature Scale (EPT-76) [6].
The CCT undertook the development of the
ITS-90 because of the deficiencies and limitations
of the IPTS-68(75) and completed the scale in ac-
cordance with Resolution 7 of the 18th Conference
Generate des Poid et Mesures [7], which met in
October 1987. The deficiencies and limitations of
the IPTS-68(75) included its lower limit of 13.81 K,
its inaccuracy relative to thermodynamic tempera-
tures, and its non-uniqueness and irreproducibility,
especially in the temperature region from
r68= 903.89 K (r68= 630.74 °C) to T58= 1337.58 K(t6S= 1064.43 °C), the region in which the Pt-
10%Rh/Pt thermocouple was the standard inter-
polating instrument.
The ITS-90 extends upward from 0.65 K and
temperatures on this scale are in much better agree-
ment with thermodynamic values than are those on
the IPTS-68(75) and the EPT-76. The new scale
has subranges and alternative definitions in certain
ranges that greatly facilitate its use. Furthermore,
its continuity, precision, and reproducibility
throughout its range are much improved over the
corresponding characteristics of the previous
scales. The replacement of the thermocouple with
the platinum resistance thermometer at tempera-
tures in the range from 630.74 to 961.93 °C of the
IPTS-68(75) resulted in the biggest improvement in
reproducibility. Also, improvements in radiometric
techniques have allowed using the silver freezing
point as the reference point for radiation ther-
mometry. This is a lower temperature reference
point than was used in the IPTS-68(75).
The change in the temperature scale affects not
only technical interests involved directly in ther-
mometry but also those involved with other refer-
ence standards, such as electrical standards, if those
standards are sensitive to temperature. As exam-
ples, standard resistors and standard cells are sensi-
tive to temperature and generally are maintained in
constant-temperature environments, at least in na-
tional standards laboratories. At the present time,
the temperatures of those environments are nor-
mally determined with thermometers that have
been calibrated on the IPTS-68(75). A given ther-
modynamic temperature expressed on the ITS-90,
however, has a value that is different from that ex-
pressed on the IPTS-68(75), as indicated [2,8] in
figure 1. A table of differences between tempera-
tures on the ITS-90, Tw or t90 , and those on the
IPTS-68(75), Tb& or tbS , and those on the EPT-76,
Tlb , is given in the text of the ITS-90. Since temper-
ature values expressed on these scales are different,
if the temperature of the environment of a refer-
ence standard is adjusted so that its value when ex-
pressed on the ITS-90 has the same value as had been
used on the IPTS-68(75), there will have been a
change of the thermodynamic temperature and the
value of the reference standard will usually change.
Of course, one may not want to change the ther-
modynamic temperature of the reference standard.
In that case, the thermodynamic temperature, as
expressed on the IPTS-68(75), can simply be ex-
pressed on the ITS-90 (a numerical value different
70
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
0.02
o00
-0.02
1
Oen -0.04
©ocow0) 4«*- *
o«b.
3<0
«Q.
E -0.2H-
1 1 r-"
1 1i
—
1
1
1
1
1
~*
•
1i
-200 /t //
/
200 f/°C
»
<
) 1 100X
i 1 I.I.I
-10mX
-20mK
\\\
1 ' 1 ..
- 0.4
- 0.2
--0.2
200 200 400
r/°C
600 800 1000
Figure 1. Differences between rTO and r6s as a function of fTO (expressed as t).
from that on the IPTS-68(75)) and the reference
standards will be unaffected. For more details on
the effects of the change of the temperature scale
on electrical standards, see National Institute of
Standards and Technology (NIST) Technical Note
1263 [4].
In addition to the effect on reference standards
for measurements, all temperature-sensitive proper-
ties that are presently expressed on the IPTS-68(75)
may be affected and may require changes in values.
For details on realizations and approximations of
the ITS-90, see NIST Technical Note 1265 [9].
2. CCT's 17th Session Principal Decisions2.1 Definition of the ITS-90
The ITS-90 was designed by the CCT in such a
manner that temperature values obtained on it donot deviate from the Kelvin thermodynamic tem-
perature values by more than the uncertainties of
the latter values at the time the ITS-90 wasadopted. Thermodynamic temperature is indicated
by the symbol T and has the unit known as the
kelvin, symbol K. The size of the kelvin is defined
to be 1/273.16 of the thermodynamic temperature
of the triple point of water.
Temperatures on the ITS-90 can be expressed in
terms of the International Kelvin Temperatures,
with the symbol T90 , or in terms of the Interna-
tional Celsius Temperatures, with the symbol t90 .
The unit of the temperature T90 is the kelvin, sym-
bol K, and the unit of the temperature t90 is the
degree Celsius, symbol °C. The relation between
T90 and t90 is:
r90/°C=7VK-273.15. (1)
The ITS-90 has alternative definitions of T90 in
certain temperature ranges and they have equal
status. In measurements of the highest precision
made at the same temperature, the alternative defi-
nitions may yield detectably different temperature
values. Also, at any given temperature between
defining fixed points, different interpolating ther-
mometers that meet the specifications of the ITS-
90 may indicate different temperature values. Themagnitude of the differences resulting from these
two sources, however, is sufficiently small to be
negligible for all practical purposes.
Temperatures on the ITS-90 are defined in terms
of equilibrium states of pure substances (defining
fixed points), interpolating instruments, and equa-
tions that relate the measured property to r90 . Thedefining equilibrium states of the pure substances
and the assigned temperatures are listed in table 1.
71
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
Table 1. Defining fixed points of the ITS-90 Table 2. Values of the coefficients A it and of the constants Aa,
B, and C for the 3He and 4He vapor pressure
temperature range for which each equation is
equations and the
Material3 Equilibrium stateb Tempierature valid
r*, (K) '90 (°C)
Coef. or
constant
3He0.65 to 3.2 K
4He1.25 to 2.1768 K
4He3He and 4He VP 3 to 5 -270.15 to 2.1768 to 5.0 K
TP 13.8033
-268.15
-259.3467e-H 2 A 1.053 447 1.392 408 3.146 631
e-H2 (or He) VP(or CVGT) = 17 = -256.15 As 0.980 106 0.527 153 1.357 655
e-H2 (or He) VP (or CVGT) = 20.3 = -252.85 A 2 0.676 380 0.166 756 0.413 923
Nec TP 24.5561 -248.5939
o 2 TP 54.3584 -218.7916 A, 0.372 692 0.050 988 0.091 159
Ar TP 83.8058 -189.3442 A* 0.151 656 0.026 514 0.016 349
Hgc TP 234.3156 -38.8344 A, -0.002 263 0.001 975 0.001 826
H 2 TP 273.16 0.01
Gac MP 302.9146 29.7646 A6 0.006 596 -0.017 976 -0.004 325
Inc FP 429.7485 156.5985 An 0.088 966 0.005 409 -0.004 973
Sn FP 505.078 231.928 A % -0.004 770 0.013 259
Zn FP 692.677 419.527
Alc FP 933.473 660.323 A, -0.054 943
Ag FP 1234.93 961.78 B 7.3 5.6 10.3
Au FP 1337.33 1064.18 C 4.3 2.9 1.9
Cu c FP 1357.77 1084.62
" e-H 2 indicates equilibrium hydrogen, that is, hydrogen with
the equilibrium distribution of its ortho and para states. Normal
hydrogen at room temperture contains 25% para hydrogen and
75% ortho hydrogen.b VP indicates vapor pressure point; CVGT indicates constant
volume gas thermometer point; TP indicates triple point (equi-
librium temperature at which the solid, liquid and vapor phases
coexist); FP indicates freezing point and MP indicates melting
point (the FP and the MP are equilibrium temperatures at which
the solid and liquid phases coexist under a pressure of 101 325
Pa, one standard atmosphere). The isotopic composition is that
naturally occurring.c Previously, these were secondary fixed points.
2.1.1 Temperature Range From 0.65 to 5.0 KBetween 0.65 and 3.2 K, the ITS-90 is defined by
the vapor pressure-temperature relation of3He, and
between 1.25 and 2.1768 K (the X point) and be-
tween 2.1768 and 5.0 K by the vapor pressure-tem-
perature relations of4He. T^ is defined by the
vapor pressure equations of the form:
value of which is determined by using either3He or
4He vapor pressure thermometry.
For a4F£e CVGT used between 4.2 K and the
triple point of neon (24.5561 K), Too is defined by
the equation:
T90=a+bp+cp :
(3)
where p is the CVGT pressure and a , b , and c are
coefficients that are determined by calibration at
the three specified temperatures, but with the addi-
tional requirement that the calibration with the
vapor pressure thermometer be made at a tempera-
ture between 4.2 and 5.0 K.
For a4He CVGT used between 3.0 and 4.2 K,
and for a3He CVGT used from 3.0 to 24.5561 K,
the non-ideality of the gas must be taken into ac-
count, using the appropriate second virial coeffi-
cient, ^(Too) or BiiTw)- 7"90 is defined in this range
by the equation:
T90/K=A Q+2 A{(\n(p/Y2L)-B)/C}\ (2)
J on—bp+cp 1
with the values of the coefficients A it and of the
constants A , B, and C of the equations being speci-
fied, and given in table 2.
2.1.2 Temperature Range From 3.0 to 24.5561 KBetween 3.0 and 24.5561 K, the ITS-90 is defined
in terms of the3He or
4He constant volume gas
thermometer (CVGT). The thermometer is cali-
brated at three temperatures—at the triple point of
neon (24.5561 K), at the triple point of equilibrium
hydrogen (see footnote a in table 1) (13.8033 K),
and at a temperature between 3.0 and 5.0 K, the
l+BAT^N/V (4)
where p is the CVGT pressure; a, b, and c are
coefficients that are determined from calibration at
the three defining temperatures; BX {T<^) refers to
B)(Tqo) or BtiTw); and N/V is the gas density in
moles per cubic meter in the CVGT bulb. The val-
ues of the second virial coefficients at any given
temperature are to be calculated according to equa-
tions specified in the official document of the ITS-
90 (and also in the NIST Technical Note 1265 [9]).
72
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
2.1.3 Temperature Range From 13.8033 to
1234.93 K Between 13.8033 K (-259.3467 °C)
and 1234.93 K (961.78 °C), the ITS-90 is defined in
terms of the specified fixed points given in table 1
,
by resistance ratios of platinum resistance ther-
mometers (PRTs) obtained by calibration at speci-
fied sets of the fixed points, and by reference
functions and deviation functions of resistance ra-
tios which relate to Tw between the fixed points.
Temperatures on the ITS-90 are expressed in
terms of the ratio W(Tw) of the resistance R (T^) at
temperature Tw and the resistance R (273.16 K) at
the triple point of water, i.e.,
W(T90)=R(T90)/R(213.16 K). (5)
For a PRT to be an acceptable instrument of the
ITS-90, its coil must be made from pure platinum
and be strain-free. Additionally, the finished PRTmust meet one of the following criteria:
W(302.9146 K)> 1.118 07
^(234.3156 K) < 0.844 235.
(6)
(7)
An acceptable PRT that is to be used to the freez-
ing point of silver must meet the following require-
ment also:
^(1234.93 K) > 4.2844. (8)
The temperature T9Q is calculated from the resis-
tance ratio relation:
W(T90)-JVt{T90)= MV(T90), (9)
where W{T<^) is the observed value, Wr(r90) is the
value calculated from the reference function, and
AW(r9o) is the deviation of the observed W{T^value of the particular PRT from the reference
function value at T^.
There are two reference functions Wr{T<^), one
for the range 13.8033 to 273.16 K and the second
for the range 273.15 to 1234.93 K. The deviation
A W^Tcio) is obtained as a function of Tn for various
ranges by calibration at specified fixed points. Theform of the deviation function depends upon the
temperature range of calibration.
2.1.4 Temperature Subrange From 13.8033 to
273.16 K In the range 13.8033 to 273.16 K, the
equation for the reference function Wt{T<^) as a
function of T^ is given by:
ln[Wt(T90)]=AQ+2 /t{[ln(7V273.16 K)1=1
+ 1.5]/1.5}'. (10)
The specified inverse of this equation, equivalent to
within ±0.000 1 K, is:
7V273.16 K=5, £ n /[ r(r90)]'/6 -0.65
0.35)'
(11)
The values of the constants A and B and values of
the coefficients Atand B
t of the two equations are
listed in table 3.
Table 3. Values of the coefficients A,, Biy C,, and D, and of the
constants A , B , Co, and A> in the reference functions, eqs (10)
and (18), and in the inverse functions approximating them, given
by eqs (11) and (19)
Constant or Value Constant or Value
coefficient coefficient
A -2.135 347 29 Bo 0.183 324 722
Ax
3.183 247 20 fi, 0.240 975 303
A 2 -1.801435 97 B2 0.209 108 771
A, 0.717 272 04 Bi 0.190 439 972
A, 0.503 440 27 B< 0.142 648 498
As -0.618 993 95 B, 0.077 993 465
A6 -0.053 323 22 B„ 0.012 475 611
A-, 0.280 213 62 Bi -0.032 267 127
At 0.107 152 24 B* -0.075 291 522
A, -0.293 028 65 5, -0.056 470 670
Aio 0.044 598 72 B\o 0.076 201 285
An 0.118 686 32 Bu 0.123 893 204
An -0.052 48134 Bn -0.029 201 193
Bn -0.091 173 542
B\A 0.001 317 696
5, 5 0.026 025 526
Ca 2.781 572 54 Do 439.932 854
c, 1.646 509 16 £>! 472.418 020
c2 -0.137 143 90 D2 37.684 494
c3 -0.006 497 67 D, 7.472 018
c4 -0.002 344 44 Z>4 2.920 828
c5 0.005 118 68 A 0.005 184
c. 0.001 879 82 A, -0.963 864
c, -0.002 044 72 D, -0.188 732
c8 -0.000 461 22 D, 0.191 203
c, 0.000 457 24 D, 0.049 025
73
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
If the PRT is to be used throughout the range
from 13.8033 to 273.16 K, it must be calibrated at
the triple points of equilibrium hydrogen (13.8033
K), neon (24.5561 K), oxygen (54.3584 K), argon
(83.8058 K), mercury (234.3156 K), and water
(273.16 K) and at two additional temperatures very
close to 17.0 and 20.3 K. The temperatures of cali-
bration at 17.0 and 20.3 K may be determined by
using either a CVGT or the vapor pressure-tem-
perature relation of equilibrium hydrogen. Whenthe CVGT is used, the two temperatures must be
within the ranges 16.9 to 17.1 K and 20.2 to 20.4 K,
respectively. When the equilibrium hydrogen va-
por pressure thermometer is used, the two temper-
atures must be within the ranges 17.025 to 17.045 Kand 20.26 to 20.28 K, respectively. The tempera-
tures of the equilibrium hydrogen vapor pressure
thermometer are determined from the values of the
hydrogen vapor pressure, p , and the equations:
7VK- 17.035=0?/kPa-33.3213)/13.32 (12)
7VK-20.27= (p/kPa- 101.292)730. (13)
Depending upon the temperature range of appli-
cation, PRTs may be calibrated from 273.16 Kdown to 13.8033 K (triple point of equilibrium hy-
drogen), down to 24.5561 K (triple point of neon),
down to 54.3584 K (triple point of oxygen), or
down to 83.8058 K (triple point of argon).
The deviation function for calibration in the
range 13.8033 to 273.16 K is given by the relation:
A^,(r90)= W(T90)- lVT(T90)=a,[W(T90)-l]
+ b][W(T90)-\]
1 +2c l[\nlV(T90)]
i+n, (14)
with n —2. The coefficients a u b u and the five c,'s
of the deviation function are obtained by calibra-
tion at all of the above eight temperatures, includ-
ing that at the triple point of water. The values of
Wr(T90) are obtained from the reference function
for this range. Although the official text of the
ITS-90 does not assign subscripts to the coefficients
a and b, nor does it designate the deviation equa-
tions by the symbols LWm{T9Q), where in eq (14)
m = 1, these designations will be used in this paper
for clarity and for ease of reference. In any case,
some such terminology must be used in PRT cali-
bration reports and this was chosen for conve-
nience.
2.1.5 Subrange From 24.5561 to 273.16 K Thedeviation function for calibration in this range is
given by the relation:
&W2(T90)= W(T90)-Wt(T90)= a 2[1V(T90)- 1]
+ b2[W(T90)-l]2+^ aWnWiTso)]'*", (15)
where the exponent n has the value n =0. The co-
efficients a 2 , b2 , and c, of this deviation function are
obtained by calibrating the PRT at the triple points
of equilibrium hydrogen (13.8033 K), neon
(24.5561 K), oxygen (54.3584 K), argon (83.8058
K), mercury (234.3156 K) and water (273.16 K).
The values of W^T^) are obtained from the refer-
ence function.
2.1.6 Subrange From 54.3584 to 273.16 K Thedeviation function for calibration in this range is
given by the relation:
A 3(r9o)=a3[^(r9o)-i]+6 3[^(r9o)-i]2
+ cl[\nW(T9o)Y
+", (16)
where the exponent n has the value n = 1. The co-
efficients a 3 , bi, and c, of this deviation function are
obtained by calibrating the PRT at the triple points
of oxygen (54.3584 K), argon (83.8058 K), mercury
(234.3156 K), and water (273.16 K). The values of
fVr(Tgo) are obtained from the reference function.
2.1.7 Subrange From 83.8058 to 273.16 K Thedeviation function for calibration in this range is
given by the relation:
MV,(T90)= a,[W(T90)-\]
+ b,[W(T90)-l]\nW(T90). (17)
The coefficients a 4 and bx of this deviation function
are obtained by calibrating the PRT at the triple
points of argon (83.8058 K), mercury (234.3156 K),
and water (273.16 K). The values of Wr{T9^) are
obtained from the reference function.
2.1.8 Temperature Subrange From 273.15 to
1234.93 K In the range 273.15 to 1234.93 K, the
equation for the reference function W,{ T<^) is given
by:
Wr(T90)=C +^Ci
7VK-48
754.15V1
)' (18)
The specified inverse of this equation, equivalent to
within ±0.000 13 K, is:
^7*0-2-64(19)
74
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
The values of the constants C and D and of the
coefficients C, and Z>, for these equations are listed
in table 3.
If the PRT is to be used over this entire subrange
(273.15 to 1234.93 K), it must be calibrated at the
triple point of water (273.16 K) and at the freezing
points of tin (505.078 K), zinc (692.677 K), alu-
minum (933.473 K), and silver (1234.93 K).
The deviation function is given by the relation:
MV6(T«>)= W(T90)- W!{T90)= a b[W{T9O)- 1]
+ b6[W(T90)-\]2+ c6[W(T90)-\]
i
+d[W(T90)-lV(933A73K)]2
. (20)
The values of the coefficients a b , b6 , and c6 are de-
termined from the measured deviations &W(T90) of
WiTgo) from the reference values Wr(T90) at the
freezing points of tin (505.078 K), zinc (692.677 K)
and aluminum (933.473 K). The coefficient d is de-
termined from these values of the coefficients a 6 , Z>6 ,
and c6 and the deviation kW(T90) of W(T90) from
the reference value WC(T90) at the freezing point of
silver. The coefficient d in this equation is used
only for those temperature measurements in the
range from the freezing point of aluminum to the
freezing point of silver. For temperature measure-
ments below the freezing point of aluminum, d =0.
PRTs may be calibrated for use over the whole
range (273.15 to 1234.93 K) or for shorter ranges
by calibrations at fixed points between 273.15 Kand the upper limit of 933.473 K (freezing point of
aluminum, 660.323 °C), of 692.677 K (freezing
point of zinc, 419.527 °C), of 505.078 K (freezing
point of tin, 231.928 °C), of 429.7485 K (freezing
point of indium, 156.5985 °C), or of 302.9146 K(melting point of gallium, 29.7646 °C).
2.1.9 Subrange From 273.15 to 933.473 K For
application in this range, the PRT is calibrated at
the triple point of water (273.16 K), and at the
freezing points of tin (505.078 K), zinc (692.677 K),
and aluminum (933.473 K). The deviation function
is given by the relation:
MV1(T90)= a 1[W(T90)-l] + b 1[W(T90)-l]2
+ Cl[W(T90)-l]\ (21)
The coefficients a-,, b-,, and c 7 , identical to a b , bb , and
c6 , respectively, are determined from the deviations
A^(r9o) of W(T90) from the reference values
WXT90) at the freezing points of tin (505.078 K),
zinc (692.677 K), and aluminum (933.473 K).
2.1.10 Subrange From 273.15 to 692.677 KFor application in this range, the PRT is calibrated
at the triple point of water (273.16 K), and at the
freezing points of tin (505.078 K) and zinc (692.677
K). The deviation function is given by the relation:
MVi(T90)= a s[W(T90)-l] + b i[W(T90)-\]2
. (22)
The coefficients a 8 and 6 8 are determined from the
deviations A ^(7^) of WiT^) from the reference
values W^Tw) at the freezing points of tin (505.078
K) and zinc (692.677 K).
2.1.11 Subrange From 273.15 to 505.078 KFor application in this range, the PRT is calibrated
at the triple point of water (273.16 K), and at the
freezing points of indium (429.7485 K) and tin
(505.078 K). The form of the deviation function is
the same as that for the subrange 273.15 to 692.677
K, i.e.,
A 9(r90)=a9[^(r90)-i]+69[^(r90)-i]2
. (23)
The coefficients a9 and b9 are determined from the
deviations A.W(T90) of W(T90) from the reference
values Wr(T90) at the freezing points of indium
(429.7485 K) and tin (505.078 K).
2.1.12 Subrange From 273.15 to 429.7485 KFor application in this range, the PRT is calibrated
at the triple point of water (273.16 K) and at the
freezing point of indium (429.7485 K). The devia-
tion function is:
A^10(r90)= aio[^(r9o)-i]. (24)
The coefficient a w is determined from the deviation
&.W(T90) of W(T9Q) from the reference value
Wr(T90) at the freezing point of indium (429.7485
K).
2.1.13 Subrange From 273.15 to 302.9146 KFor application in this range, the PRT is calibrated
at the triple point of water (273.16 K) and at the
melting point of gallium (302.9146 K). The devia-
tion function is:
MVn(T90)=an[W(T90)-l]. (25)
The coefficient a n is determined from the deviation
AJf(r90) of W(T90) from the reference value
Wr(Tgo) at the melting point of gallium (302.9146
K).
2.1.14 Subrange From 234.3156 to 302.9146 KFor application in this range, the PRT is calibrated
at the triple points of mercury (234.3156 K) and
water (273.16 K), and at the melting point of gal-
75
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
Hum (302.9146 K). The form of the deviation func-
tion is the same as that for the subrange 273.15 to
692.677 K, i.e.,
^W5(T90)=a5[W(T90)-l]+ bs[W(T90)-l]2
. (26)
The coefficients a 5 and b5 are determined from the
deviations LW{T^ of W{T<^) from the reference
values Wr(Tgo) at the triple point of mercury
(234.3156 K) and at the melting point of gallium
(302.9146 K). The reference values Wt(T^) must be
calculated from the relevant reference function,
both reference functions being required to cover
this range.
2.1.15 Temperature Range Above 1234.93 KAt temperatures above 1234.93 K, T90 is defined by
the relation:
L x(Tn) _ exp[c2/\rqo(X)]- 1
(27)
in which Lx(7ro) and Lx[T90(X)] are the spectral
concentrations of the radiance of a blackbody at
wavelength X (in vacuum) at T^ and at T^X), re-
spectively. r^X) refers to either the silver freezing
point [T90(Ag)= 1234.93 K], the gold freezing point
[7'9 (Au)= 1337.33 K] or the copper freezing point
[r9o(Cu)= 1357.77 K]. c 2= 0.014388 m-K. Although
the freezing-point temperature of silver is the junc-
tion point of platinum resistance thermometry and
radiation thermometry, it is believed that the Twvalues of the freezing points of silver, gold and
copper are sufficiently self-consistent that the use
of any one of them as the reference temperature
F9o(X) will not result in any significant difference in
the measured values of T90 from what would be
obtained if only the silver freezing point were used.
2.2 Recommendations of the CCT
Three recommendations were adopted by the
CCT at its 17th Session. These recommendations
were considered by the CIPM and Recommenda-tion Tl (1989) of the CCT was adopted as
Recommendation 5 (CI-89) of the CIPM. Recom-mendations T2 (1989) and T3 (1989) of the CCTwere noted by the CIPM as CCT recommenda-tions. The CCT recommendations were as follows:
Recommendation Tl (1989)
The International Temperature Scale of 1990
The Comite Consultatif de Thermometrie (CCT)acting in accordance with Resolution 7 of the 18
e
CGPM has generated the International Tempera-
ture Scale of 1990 (ITS-90) in order to supersede
the International Practical Temperature Scale of
1968 (IPTS-68).
The CCT notes that, by comparison with the
IPTS-68, the ITS-90
—extends to lower temperatures, down to 0.65
K, and hence also supersedes the EPT-76,
—is in substantially better agreement with corre-
sponding thermodynamic temperatures,
—has much improved continuity, precision, and
reproducibility throughout its range and
—has subranges and alternative definitions in
certain ranges which greatly facilitate its use.
The CCT also notes that, to accompany the text
of the ITS-90 there will be two further documents,
the Supplementary Information for the ITS-90 and
Techniques for Approximating the ITS-90. These
documents will be published by the BIPM and pe-
riodically updated.
The CCT recommends
—that on January 1, 1990 the ITS-90 come into
force and
—that from this same date the IPTS-68 and the
EPT-76 be abrogated.
Recommendation T2 (1989)
Reference Tables for Thermocouples and Industrial
Platinum Resistance Thermometers
The Comite Consultatif de Thermometrie,
considering
—that the introduction of the International Tem-perature Scale of 1990 (ITS-90) will lead to an ur-
gent requirement for new reference tables for both
thermocouples and industrial platinum resistance
thermometers,
requests its Working Group 2
—to collaborate with national laboratories in the
rapid preparation of new reference tables taking
into account not only the change from IPTS-68 to
ITS-90 but also new information on the behavior of
thermocouples and industrial platinum resistance
thermometers,
recommends
—that these new tables be used as the basis for
new national and international reference tables for
thermocouples and industrial platinum resistance
thermometers and
—that meanwhile the existing reference tables
based upon IPTS-68 should be used in conjunction
with the table of differences Tw— TbS which ap-
pears in the ITS-90.
76
Volume 95, Number 1, January-February 1990
Journal of Research of the National Institute of Standards and Technology
(Note: the table of differences T^— Tbi referred to
here may be obtained also from NIST from the
author of this article).
Recommendation T3 (1989)
The Uncertainty Inherent in the Realization of the
International Temperature Scale of 1990
The Comite Consultatif de Thermometrie,
considering the requirement for assigning an un-
certainty to the numerical value of any temperature
on the International Temperature Scale of 1990
(ITS-90),
encourages national laboratories to
a) quantify the uncertainties in the fixed point
realizations,
b) quantify the uncertainties resulting from the
use of the specified interpolating instruments of
ITS-90,
c) develop the mathematical procedures describ-
ing the propagation of these uncertainties to any
intermediate temperature.
3. Conclusion
About the author: B. W. Mangum, a physicist, is a
Group Leader in the Temperature and Pressure Divi-
sion of the NIST Center for Chemical Technology
and is the NIST representative to the Comite Consul-
tatif de Thermometrie.
4. References
[1] BIPM Com. Cons. Thermometrie 17, 1989, in press.
[2] The International Temperature Scale of 1990, Metrologia
27, 3 (1990).
[3] Proces-Verbaux des seances du Comite International des
Poids et Mesures, (78=session, octobre 1989), in press.
[4] Belecki, N. B., Dziuba, R. F., Field, B. F., and Taylor,
B. N., Guidelines for Implementing the New Representa-
tions of the Volt and Ohm Effective January 1, 1990, NISTTechnical Note 1263 (June 1989).
[5] The International Temperature Scale of 1968, AmendedEdition of 1975, Metrologia 12, 7 (1976).
[6] The 1976 Provisional 0.5 K to 30 K Temperature Scale,
Metrologia 15, 65 (1979).
[7] Comptes rendus des seances de la dix-huitieme Conference
Generate des Poids et Mesures, Resolution 7 (1987).
[8] Quinn, T. J., Metrologia 26, 69 (1989).
[9] Mangum, B. W., and Furukawa, G. T., Guidelines for Re-
alizing the International Temperature Scale of 1990, NISTTechnical Note 1265 (1990), in press.
Although the uncertainties in the values of ther-
modynamic temperatures above 100 °C used in the
definition of the ITS-90 were larger than desired
and larger than had been anticipated a few years
ago, the agreement of temperatures on the ITS-90
with thermodynamic temperatures is nevertheless a
significant improvement over that of previous
scales. The replacement of the thermocouple with
the platinum resistance thermometer as the stan-
dard instrument of the scale at temperatures in the
IPTS-68(75) range from 630.74 to 961.93 °C has
improved the reproducibility over that of the
IPTS-68(75) significantly. Also, advances in radio-
metric techniques have improved the precision of
measurements in radiation thermometry. The pre-
cision of the scale, or what has been called the non-
uniqueness of the scale, is significantly improved
over that of the IPTS-68(75), as is also the scale's
continuity. The extension of the scale downward in
temperature to 0.65 K and the use of subranges
over which thermometers may be calibrated makethe ITS-90 more useful and much more flexible
than were the previous scales.
77
6.2 In this appendix, excerpts from proceedings of some meetings of the CIPMand of the CGPM that are related to the designation of the size of the kelvinand to the ITS -90 are presented.
6.2.1 The following is an excerpt from Proces-Verbaux des Seances, DeuxiemeSeries, Tome XXIV, pp. 81-82, Session de 1954 du Comite International des Poidset Mesures, published by Gauthier Villars, Editeur-Imprimeur-Libraire, Paris,
1955, concerning the recommendation of the CIPM that the numerical value of273.16 °K, exactly, be assigned to the triple point of water on the thermodynamicscale
.
"Le Comite International des Poids et Mesures recommande que l'on ddfinissedesormais 1'echelle thermodynamique au moyen du point triple de l'eau comme pointfixe fondamental .
"
"II recommande que l'on attribue par definition A la temperature du pointtriple de l'eau dans 1'echelle thermodynamique la valeur numerique 273 ,16°
K
exactement .
"
6.2.2 The following is an excerpt from Comptes Rendus des Seances, DixiemeConference Generale des Poids et Mesures, Reunie a Paris en 1954, p. 79, publishedby Gauthier Villars, Editeur-Imprimeur-Libraire, Editeur du Bureau Internationaldes Poids et Mesures, Paris, 1955, concerning the definition of the thermodynamicscale of temperature. This entails the assignment of the value 273.16 °K,
exactly, to the triple point of water as a fundamental fixed point.
DEFINITION DE L'ECHELLE THERMODYNAMIQUE DE TEMPERATURE.
RESOLUTION 3
"La Dixieme Conference Generale des Poids et Mesures decide de definir1'echelle thermodynamique de temperature au moyen du point triple de l'eau commepoint fixe fondamental, en lui attribuant la temperature 273,16 degres Kelvin,exac tement .
"
118
6.2.3 The following recommendations of the CCT, concerning the development ofthe ITS-90 and simplified methods of its realization, that were presented to the
CIPM on 9 July 1987 were excerpted from Proces-Verbaux des Seances du ComiteInternational des Poids et Mesures, Proces-Verbaux de la 76 e session - 1987, Tome
55, pp. T7-T8, Edite par le BIPM, Pavilion de Breteuil, F-92312 Sevres Cedex,France
.
9 juillet 1987
Recommandat ionsdu Comite Consultatif de Thermometrie
presenteesau Comite International des Poids et Mesures
Necessite de travaux fondamentaux en thermometrie
RECOMMANDATION T 1 (1987)
Le Comite Consultatif de Thermometrie,
tout en dSclarant que la valeur T90 de la temperature mesuree dans l'Echelleinternationale de temperature de 1990 (EIT-90) sera aussi voisine de la
temperature thermodynamique T que le permettent les connaissances actuelles,
reconnalt
— que 1' adoption de 1' EIT-90 ne signifie en aucune facon que nous ayonsmaintenant une connaissance definitive de la valeur des temperaturesthermodynamiques
,
— que le travail necessaire pour developper cette connaissance fondamentalepar des determinations directes des temperatures thermodynamiques est difficile,complexe et de longue haleine,
— qu'il est necessaire d'anticiper les besoins a venir d' exactitude etde precision croissantes,
propose, en consequence, de surveiller en permanence 1' accord entre T90 etT au fur et a mesure des progres de la science,
insiste sur 1' importance de 1' etude et de la mise en uvre de methodesnouvelles conduisant a des determinations plus exactes des temperaturesthermodynamiques, dans tout le domaine couvert par 1' EIT-90 et aussi vers les
tres basses temperatures jusqu'a quelques millikelvins
,
et insiste sur 1' importance de la poursuite des travaux des laboratoiresnationaux dans ce but.
119
Methodes simplifiees et economiques de mesure des temperatures
RECOMMANDATION T 2 (1987)
Le Comite Consultatif de Thermometrie,
tout en reconnaissant que sa tache princlpale est de veiller a ameliorerla misc en pratique et a mettre a jour periodiquement l'Echelle internationalede temperature, s'est efforce dans les annees recentes de repondre aux demandesdes chercheurs , ingenieurs et autres utllisateurs qui ont besoin de bonnesmesures de temperature plutot que de mesures de temperature du niveau le pluseleve,
a note-
les travaux recents sur les thermocouples or/platine etpalladium/platine indiquant que ces thermocouples peuvent presenter des avantagesappreciables
,pour de nombreux usages, par rapport aux thermocouples habituels
platine/platine rhodie.
En consequence, le Comite Consultatif de Thermometrie,
reconunande aux laboratoires nationaux de poursuivre leurs recherches sur
les methodes simplifiees et de cout reduit de mesure des temperatures, enparticulier sur ces nouveaux thermocouples pour determiner leur fonction ainsique les regies pratiques de construction et d' utilisation qui devraient permettred' exploiter au mieux leurs possibilites
.
120
6.2.4 The following resolution, instructing the CIPM to develop the ITS -90,
was adopted by the 18th CGPM in 1987, and is excerpted from Comptes Rendus , 18 e
Conference Generale des Poids et Mesures (1987), p. 101, Edite par le BIPM,Pavilion de Breteuil, F-92312 Sevres Cedex, France.
Mise au point d'une nouvelle echelle international de temperature
RESOLUTION 7
La Dix-huitieme Conference Generale des Poids et Mesures,
considerant
que 1'uniformite mondiale et la stabilite a long terme des mesures detemperature sont d'une importance majeure pour la science, le commerce etl'industrie, du point de vue technique comme du point' de vue economique,
etroque la valeur de la temperature dans 1' echelle pratique doit correspondreitement a la valeur de la temperature thermodynamique dans le SI,
que la Quinzieme Conference Generale des Poids et Mesures, ayant prisconnaissance des ecarts qui ont ete mis en evidence entre la valeur de latemperature dans 1' Echelle internationale pratique de temperature de 1968(EIPT-68) et la valeur correspondante de la temperature thermodynamique, a chargele Comite International des Poids et Mesures d'etablir une nouvelle echelle quidevrait etre sensiblement plus exacte,
qu'une nouvelle echelle pourra etre proposee en 1989, echelle telle queles ecarts residuels avec la temperature thermodynamique seront negligeablespendant de nombreuses annees a venir,
invite le Comite International des Poids et Mesures et les laboratoiresnationaux a achever la mise au point de cette nouvelle echelle de temperatureet a fixer une date pour sa mise en application simultanement dans tous les pays;
la valeur des ecarts entre cette nouvelle echelle et 1' EIPT-68 devrait etreannoncee au mo ins un an avant 1' adoption de la nouvelle echelle qui pourraitintervenir au ler Janvier 1990.
121
6.3 We present here examples of some typical Report of Calibration documentsfrom the NIST Calibration Laboratories for thermometers calibrated over variousrange
.
6.3.1 The following report is an example of a Report of Calibration for an SPRTcalibrated over the range from 13.8033 K to 505.078 K.
122
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
REPORT OF CALIBRATIONInternational Temperature Scale of 1990
Platinum Resistance ThermometerLAN Model 8164
Serial Number 1812284
Submitted by
National Institute of Standards and Technology
Gaithersburg, Maryland
This thermometer was calibrated with an AC bridge at a frequency of 30 Hz using a continuous measuring current
of 1.0 mA In accordance with the International Temperature Scale of 1990 (ITS-90) that was officially adopted by
the Comite* International des Poids et Mesures (CIPM) in September 1989, the subranges from 13.8033 K to 273.16 Kand 273.15 K to 505.078 K, with the following fixed points and their stated uncertainties, were used to calibrate the
thermometer:
Fixed Point
H2 TP17 K VP20 K BPNe TP2 TPAr TPHg TPH2 TPIn FPSn FP
The following values were determined for the coefficients of the pertinent deviation functions of the ITS-90, as given
in the attached material describing the scale. The attached tables were generated using these values.
Coefficients for 1.0 mA Calibration
a, = -2.0257300E-04 b, = -2.7691 191E-05c, - 1.3443513E-05 ^ = 5.9700519E-06c, = 1.1044359E-06 c, = 9.7199229E-08c5
= 3.3585947E-09a, = -2.4194948E-04 b, = -2.3366736E-05
The resistance of this thermometer at 273.16 K was calculated to be 25.4975 ohms at 1.0 mA During calibration,
the resistance at 273.16 K changed by the equivalent of 0.1 mK at 1.0 mA.
This thermometer is satisfactory as a defining instrument of the ITS-90 in accordance with the criteria that
W(302.9146 K) fc 1.11807 or W(234.3156 K) < 0.844235.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 456123
Chief, Chemical Process Metrology Division 1 January 1990
Center for Chemical Technology P.O. No. 123-456
Temperature Uncertainty
T„(K) »90(°C) (mK)
13.8033 -259.3467 ±0.217.0357 -256.1143 ±0.220.2711 -252.8799 ±0.224.5561 -248.5939 ±0.254.3584 -218.7916 ±0.183.8058 -189.3442 ±0.1
234.3156 -38.8344 ±0.1273.16 0.01 ±0.1429.7485 156.5985 ±0.7505.078 231.928 ±1.0
1 January 1990 ITS-90 Table for PRT S/N 1812284 at mA
T(kelvin) W(T) dT/dW(T) T(kelvin)
13,.8 0..0011964213,.9 0..0012209114,.0 0..0012459514..1 0.,0012715514,.2 0.,0012977114,.3 0.,0013244514,.4 0,,0013517814,,5 0.,0013797014,.6 0.,0014082214,.7 0.,0014373514.,8 0.,0014670914.,9 0.,0014974715.,0 0. 00152847
4083039934390633821737394,36595,35818,35063,34330,33617,32924,32250,
101422736450955625832944332370443944
151515151515151515.815.916161616161616.616.716.816.917171717171717,
17,
17.817.91818181818181818,
18.818.91919191919.419.519.619.719.819.920.0
0,
0,
0,
0,
0,
0,
0,
0.
0.
0.
0.
0.
0.
0,
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.0.0.
0.0.0.0.
0.0.0.
0.
0.0.
0.
0.
0.0.
0.0.
0.
0.
0.
0.
0.0.
0.
W(T)
,00152847,00156012,00159242,00162538,00165901,00169331,00172829,00176396,00180033001837410018752000191372001952960019929500203367002075150021173900216040002204180022487500229410002340260023872200243499002483580025330000258325002634340026862800273907002792730028472600290266002958950030161200307419003133160031930500325385003315570033782200344180003506330035718000363823003705620037739800384330003913610039849000405718
dT/dW(T)
31596 .27630959 .85730341 .11729739 .50229154 .47128585 .49928032 .07627493 .70826969 .91426460 .23225964 .21225481 .41925011 .43424553 .85024108 .27623674 .33123251 .65222839 .88422438 .68622047,.73021666,.69821295,.28220933,.18820580,.12920235,.83019900,.02519572,.45719252.,87918941.,04918636.,73918339.,72518049.,79017766. 72817490. 33717220. 42316956. 79916699. 28316447. 70016201. 88215961. 66415726. 88815497. 40215273. 05815053. 71214839. 22714629. 46814424. 30614223. 61614027. 27613835. 170
1 January 1990 ITS-90 Table for PRT S/N 1812284 at 0mA
T(kelvin)
20.020.120.220.320.420.520.620.720.820.921.021.121.221.321.421.521.621.721.821.922.02222222222222222.822.9232323232323232323.823.924.02424242424242424.824.925.0
W(T)
0.004057180.004130450.004204730.004280010.004356310.004433620.004511960.004591330.004671730.004753170.004835650.004919180.005003770.005089420.005176130.005263910.005352760.005442700.005533710.005625810.005719010.005813300.005908690.006005190.006102790.006201510.006301340.006402300.006504380.006607580.006711920.006817400.006924010.007031770.007140670.007250720.007361930.007474290.007587800.007702480.007818330.007935340.008053520.008172880.008293410.008415110.008538000.008662070.008787330.008913770.00904141
dT/dW(T) T(kelvin)
13647 .18313463 .20513283 .13013106 .85512934 .27912765 .30612599 .84212437 .79612279 .08012123 .60711971 .29611822 .06611675 .83911532 .53911392 .09411254 .43111119 .48210987 .17910857..45910730..25710605..51210483,.16510363,.15710245,.43410129,,94010016.,6229905.,4289796.,3109689.,2179584.,1039480.,9229379.,6299280. 1819182. 5359086. 6508992. 4878900. 0068809. 1698719. 9418632. 2848546. 1658461. 5488378. 4018296. 6938216. 3918137. 4648059. 8857983. 6227908. 6487834. 937
25252525.
2525,25.
25,25.825.9262626262626262626.826.927.02727272727272727.827.928.0282828282828.628.728.828.929.029.1292929292929,
29.829.930.0
W(T)
0.009041410.009170230.009300250.009431470.009563880.009697490.009832300.009968320.010105540.010243960.010383590.010524440.010666490.010809750.010954230.011099910.011246820.011394940.011544270.011694820.011846590.011999580.012153790.012309220.012465870.012623730.012782820.012943140.013104670.013267420.013431400.013596600.013763020.013930660.014099520.014269600.014440910.014613440.014787180.014962150.015138330.015315740.015494360.015674200.015855260.016037530.016221020.016405720.016591630.016778760.01696710
dT/dW(T)
7762 .4607691 .1917621 .1067552 .1807484 .3887417 .7077352 .1137287 .5857224 .1017161 .6397100 .1787039 .6996980 .1816921 .6056863 .9536807 .2056751,.3456696,.3546642,.2166588,.9136536,.4306484,.7506433,,8586383,.7396334,.3796285,,7616237.,8746190..7026144.,2316098.,4506053.,3456008.,9045965. 1135921. 9625879. 4385837. 5305796. 2265755. 5165715. 3895675. 8345636. 8415598. 4005560. 5015523. 1355486. 2925449. 9625414. 1385378. 8095343. 9685309. 606
6.3.2 The following report is an example of a Report of Calibration for an SPRT
calibrated over the range from 83.8058 K to 933.473 K.
126
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAJTHERSBURG, MARYLAND 20899
REPORT OF CALIBRATIONInternational Temperature Scale of 1990
Platinum Resistance ThermometerChino Model R800-2
Serial Number RS59A-8
Submitted by
National Institute of Standards and Technology
Gaithersburg, Maryland
This platinum resistance thermometer (PRT) was calibrated with an AC bridge at a frequency of 30 Hz using
continuous measuring currents of 1.0 mA and 2.0 mA. In accordance with the International Temperature Scale
of 1990 (ITS-90) that was officially adopted by the Comity International des Poids et Mesures (CIPM) in
September 1989, the subranges from 83.8058 K to 273.16 K and 273.15 K to 933.473 K, with the following fixed
points and their stated uncertainties, were used to calibrate the thermometer:
Fixed Point
Ar TPHg TPH2 TPSn FPZn FPAl FP
Temperature Uncertainty
T,o(K) t*>(°C) (mK)
83.8058 -189.3442 ±0.1234.3156 -38.8344 ±0.1273.16 0.01 ±0.1505.078 231.928 ±1.0692.677 419.527 ±1.0933.473 660.323 ±1.0
The following values were determined for the coefficients of the pertinent deviation functions of the ITS-90, as
given in the attached material describing the scale. The attached tables were generated using these values.
Coefficients for Zero-Power Dissipation Calibration Coefficients for 1.0 mA Calibration
a< = -3.2324853E-04 b, = 1.2626274E-05 a4= -3.2877518E-04 b4
= 1.1459514E-05a7
= -3.64615 15E-04 b7 = -8.5363999E-06 a7= -3.5921484E-04 b7
= -1.8940305E-05c, = 1.4664695E-06 c, = 3.7262747E-06
The resistance of this thermometer at 273.16 K was calculated to be 25.0694 ohms at zero-power dissipation and25.0696 ohms at 1.0 mA During calibration, the resistance at 273.16 K changed by the equivalent of 0.4 mKat zero-power dissipation and 0.3 mK at 1.0 mA
This thermometer is satisfactory as a defining instrument of the ITS-90 in accordance with the criteria that
W(302.9146 K) > 1.11807 or W(234.3156 K) < 0.844235. For the PRT to be acceptable for use between the
freezing points of aluminum and silver, it must also meet the additional criterion: W( 1234.93 K) > 4.2844.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 456123Chief, Chemical Process Metrology Division 1 January 1990Center for Chemical Technology P.O. No. 123-456
1 January 1990 ITS-90 Table for PRT S/N RS59A-8 at mA
t (Celsius)
1
2
3
4
56
7
89
1011121314151617181920212223242526272829303132333435363738394041424344454647484950
W(t)
0.999960121.003946601.007931861.011915901.015898721.019880321.023860701.027839871.031817821.035794561.039770081.043744381.047717471.051689341.055660001.059629451.063597681.067564701.071530511.075495101.079458481.083420661.087381621.091341371.095299901.099257231.103213351.107168261.111121961.115074461.119025741.122975821.126924681.130872341.134818801.138764041.142708081.146650921.150592541.154532961.158472181.162410191.166347001.170282601.174216991.178150191.182082181.186012961.189942541.193870921.19779810
dt/dW(t)
250.8476250.9248251.0016251.0784251.1552251.2320251.3089251.3857251.4626251.5395251.6164251.6934251.7703251.8473251.9243252.0014252.0784252.1555252.2326252.3097252.3869252.4640252.5412252.6184252.6957252.7730252.8503252.9276253.0049253.0823253.1597253.2371253.3146253.3921253.4696253.5472253.6247253.7023253.7800253.8576253.9353254.0130254.0908254.1686254.2464254.3242254.4021254.4800254.5579254.6359
t (Celsius)
5051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899
100
W(t)
1.197798101.201724071.205648841.209572411.213494781.217415951.221335911.225254681.229172241.233088601.237003761.240917721.244830491.248742051.252652411.256561571.260469541.264376301.268281871.272186241.276089411.279991381.283892161.287791731.291690111.295587301.299483281.303378071.307271661.311164061.315055261.318945271.322834081.326721691.330608111.334493331.338377361.342260201.346141831.350022281.353901531.357779591.361656451.365532131.369406601.373279891.377151981.381022881.384892591.388761101.39262843
dt/dW(t)
254.7139254.7919254.8700254.9481255.0262255.1043255.1825255.2607255.3390255.4173255.4956255.5740255.6523255.7308255.8092255.8877255.9662256.0448256.1234256.2020256.2806256.3593256.4381256.5168256.5956256.6744256.7533256.8322256.9111256.9901257.0691257.1481257.2272257.3063257.3854257.4646257.5438257.6231257.7024257.7817257.8610257.9404258.0198258.0993258.1788258.2583258.3379258.4175258.4971258.5768
6.3.3 The following report is an example of a Report of Calibration for an SPRTcalibrated over the range from 273.15 K to 1234.93 K.
129
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
REPORT OF CALIBRATIONInternational Temperature Scale of 1990
Platinum Resistance ThermometerChino Model R800-3, 0.25 ohm
Serial Number RS78A-2
Submitted by
National Institute of Standards and Technology
Gaithersburg, Maryland
This platinum resistance thermometer (PRT) was calibrated with an AC bridge at a frequency of 30 Hz using
continuous measuring currents of 5.0 mA and 10.0 mA In accordance with the International Temperature Scale
of 1990 (ITS-90) that was officially adopted by the Comite" International des Poids et Mesures (CIPM) in
September 1989, the subrange from 273.15 K to 1234.93 K, with the following fixed points and their stated
uncertainties, was used to calibrate the thermometer:
Fixed Point Temperature Uncertainty
TW(K) t*,(°C) (mK)
H2 TP 273.16 0.01 ±0.1Sn FP 505.078 231.928 ±1.0Zn FP 692.677 419.527 ±1.0Al FP 933.473 660.323 ±1.0Ag FP 1234.93 961.78 ±2.0
The following values were determined for the coefficients of the pertinent deviation functions of the ITS-90, as
given in the attached material describing the scale. The attached tables were generated using these values.
Coefficients for Zero-Power Dissipation Calibration Coefficients for 5.0 mA Calibration
a6 = -1.1296072E-04 b6= 1.1080496E-04 a6
= -9.2794623E-05 b6= 8.3014954E-05
c6 = -3.5516098E-05 d = 3.6725603E-04 c6 = -2.7075707E-05 d = 2.8767080E-04
The resistance of this thermometer at 273.16 K was calculated to be 0.2553 ohms at zero-power dissipation and0.2553 ohms at 5.0 mA During calibration, the resistance at 273.16 K changed by the equivalent of 2.0 mK at
zero-power dissipation and 1.6 mK at 1.0 mA
This thermometer is satisfactory as a defining instrument of the ITS-90 in accordance with the criteria that
W(302.9146 K) > 1.11807 or W(234.3156 K) < 0.844235. For the PRT to be acceptable for use between the
freezing points of aluminum and silver, it must also meet the additional criterion: W(1234.93 K) > 4.2844.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 456123Chief, Chemical Process Metrology Division 1 January 1990
Center for Chemical Technology P.O. No. 123-456
1 January 1990 ITS-90 Table for PRT S/N RS78A-2 at mA
t(Celsius)
1
2
3
4
56
7
8
9
1011121314151617181920212223242526272829303132333435363738394041424344454647484950
W(t)
0.999960111.003947601.007933861.011918911.015902751.019885371.023866781.027846971.031825951.035803721.039780271.043755611.047729741.051702661.055674371.059644871.063614161.067582231.071549101.075514761.079479211.083442461.087404491.091365321.095324941.099283351.103240561.107196561.111151351.115104941.119057321.123008501.126958471.130907241.134854801.138801161.142746311.146690271.150633021.154574561.158514911.162454051.166391991.170328721.174264261.178198591.182131731.186063661.189994391.193923921.19785226
dt/dW(t)
250.7845250.8613250.9378251.0144251.0909251.1675251.2441251.3207251.3973251.4740251.5507251.6274251.7041251.7808251.8576251.9344252.0112252.0880252.1649252.2417252.3186252.3956252.4725252.5495252.6265252.7036252.7806252.8577252.9348253.0120253.0891253.1663253.2436253.3208253.3981253.4754253.5528253.6301253.7075253.7850253.8624253.9399254.0175254.0950254.1726254.2502254.3279254.4055254.4833254.5610
t(Celsius)
5051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899
100
W(t)
1.197852261.201779391.205705321.209630051.213553581.217475921.221397051.225316991.229235731.233153271.237069611.240984761.244898711.248811461.252723011.256633371.260542521.264450491.268357261.272262831.276167201.280070381.283972371.287873161.291772751.295671151.299568351.303464361.307359181.311252801.315145231.319036471.322926511.326815361.330703011.334589471.338474741.342358821.346241701.350123401.354003901.357883211.361761321.365638251.369513981.373388531.377261881.381134041.385005011.388874801.39274339
dt/dW(t)
254.6388254.7166254.7944254.8723254.9502255.0282255.1061255.1842255.2622255.3403255.4184255.4965255.5747255.6529255.7312255.8094255.8878255.9661256.0445256.1229256.2014256.2799256.3584256.4369256.5155256.5942256.6728256.7515256.8303256.9090256.9878257.0667257.1456257.2245257.3034257.3824257.4614257.5405257.6196257.6987257.7779257.8571257.9363258.0156258.0949258.1743258.2536258.3331258.4125258.4920
6.3.4 The following is an example of a Report giving the T68 to Tgo conversionfor a long- stem type SPRT calibrated on the IPTS-68(75) from the oxygencondensation point to the freezing point of zinc.
132
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
TEST REPORTInternational Temperature Scale of 1990
Platinum Resistance Thermometer
Chino Model R800-2
Serial Number RS8YA-5
Submitted by
National Institute of Standards and Technology
Gaithersburg, Maryland
This thermometer was calibrated 1 March 1989 on the International Practical Temperature Scale of 1968 (IPTS-68)
and that calibration has been converted to the International Temperature Scale of 1990 (ITS-90). The IPTS-68calibration was performed with a DC bridge at a continuous current of 1.0 mA The WfT
1
,,,) values for the triple
point of mercury and the triple point of argon were calculated from the IPTS-68 coefficients.
Fixed Point Temperature Uncertainty
T«,(K) t»(°C) (mK)
Ar TP 83.8058 -189.3442 ±0.1
Hg TP 234.3156 -38.8344 ±0.1H2 TP 273.16 0.01 ±0.1
Sn FP 505.078 231.928 ±1.0Zn FP 692.677 419.527 ±1.0
The following values were determined for the coefficients of the pertinent deviation functions of the ITS-90, as given
in the attached material describing the scale. The attached tables were generated using these values.
Coefficients for 1.0 mA Calibration
a 4= -9.3225823E-05 b< = -9.9914440E-06
a 8= -9.1058813E-05 b8
= -7.6061559E-06
The resistance of this thermometer at 273.16 K was calculated to be 25.5096 ohms at 1.0 mA. During calibration,
the resistance at 273.16 K changed by the equivalent of 0.5 mK at 1.0 mA
This thermometer is satisfactory as a defining instrument of the ITS-90 in accordance with the criteria that
W(302.9146 K) > 1.11807 or W(234.3156 K) < 0.844235. For the PRT to be acceptable for use between the freezing
points of aluminum and silver, it must also meet the additional criterion: W(1234.93 K) > 4.2844.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 456123Chief, Chemical Process Metrology Division 1 January 1990Center for Chemical Technology P.O. No. 123-456
1 January 1990 ITS-90 Table for PRT S/N RS8YA-5 at 1mA
t(Celsius)
1
2
3
4
56
7
8
9
1011121314151617181920212223242526272829303132333435363738394041424344454647484950
W(t)
0.999960111.003947681.007934031.011919161.015903071.019885761.023867231.027847491.031826531.035804351.039780961.043756351.047730521.051703481.055675231.059645761.063615081.067583181.071550081.075515761.079480221.083443481.087405521.091366361.095325981.099284391.103241591.107197591.111152371.115105941.119058311.123009461.126959411.130908151.134855691.138802011.142747131.146691041.150633751.154575251.158515541.162454631.166392521.170329191.174264671.178198941.182132001.186063861.189994521.193923981.19785223
dt/dW(t)
250.7791250.8562250.9329251.0097251.0865251.1633251.2401251.3170251.3938251.4707251.5476251.6245251.7015251.7784251.8554251.9324252.0095252.0865252.1636252.2407252.3178252.3950252.4721252.5493252.6266252.7038252.7811252.8584252.9357253.0131253.0904253.1679253.2453253.3228253.4003253.4778253.5553253.6329253.7105253.7882253.8658253.9435254.0212254.0990254.1768254.2546254.3325254.4103254.4883254.5662
t (Celsius)
5051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899
100
W(t)
1.197852231.201779281.205705121.209629771.213553211.217475451.221396481.225316321.229234961.233152391.237068621.240983661.244897491.248810121.252721551.256631791.260540821.264448661.268355291.272260731.276164971.280068011.283969851.287870501.291769941.295668191.299565241.303461101.307355761.311249221.315141491.319032551.322922431.326811111.330698591.334584881.338469971.342353861.346236571.350118071.353998391.357877511.361755431.365632161.369507701.373382051.377255201.381127161.384997921.388867501.39273588
dt/dW(t)
254.6442254.7222254.8002254.8783254.9564255.0345255.1127255.1909255.2691255.3474255.4257255.5040255.5824255.6608255.7392255.8177255.8962255.9747256.0533256.1319256.2105256.2892256.3679256.4466256.5254256.6042256.6830256.7619256.8408256.9198256.9988257.0778257.1568257.2359257.3150257.3942257.4734257.5526257.6319257.7112257.7905257.8698257.9493258.0287258.1082258.1877258.2672258.3468258.4264258.5061
6.3.5 The following is an example of a Report giving the T68 to r90 conversionfor a capsule -type SPRT calibrated on the IPTS-68(75) from the triple point ofequilibrium hydrogen to the freezing point of tin.
135
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
TEST REPORTInternational Temperature Scale of 1990
Platinum Resistance ThermometerLAN Model 8164
Serial Number 1812284
Submitted by
National Institute of Standards and Technology
Gaithersburg, Maryland
This thermometer was calibrated 21 September 1989 on the International Practical Temperature Scale of 1968
(IPTS-68) and that calibration has been converted to the International Temperature Scale of 1990 (ITS-90). TheIPTS-68 calibration was performed with an AC bridge at a continuous current of 1.0 mA. The W(TW) values for the
triple point of neon, the triple point of mercury, the triple point of argon, and the freezing point of indium werecalculated from the IPTS-68 coefficients.
Fixed Point Temperature Uncertainty
T»(K) t»(°C) (mK)
H2 TP 13.8033 -259.3467 ±0.217 K VP 17.0301 -256.1143 ±0.220 K BP 20.2701 -252.8799 ±0.2Ne TP 24.5561 -248.5939 ±0.2o2 TP 54.3584 -218.7916 ±0.2At TP 83.8058 -189.3442 ±0.1Hg TP 234.3156 -38.8344 ±0.1H2 TP 273.16 0.01 ±0.1In FP 429.7485 156.5985 ±0.7Sn FP 505.078 231.928 ±1.0
The following values were determined for the coefficients of the pertinent deviation functions of the ITS-90, as given
in the attached material describing the scale. The attached tables were generated using these values.
Coefficients for 1.0 mA Calibration
a, = -2.5239001 E-04 b, = -1.2277862E-04c, = -2.3783015E-06 Cj = -4.3892024E-06
Cj = -1.5608728E-06 c4= -2.1374663E-07
c5= -1.0344171E-08
a, = -2.5287142E-04 b, = -1.1130131E-05
The resistance of this thermometer at 273.16 K was calculated to be 25.4975 ohms at 1.0 mA During calibration,
the resistance at 273.16 K changed by the equivalent of 0.1 mK at 1.0 mA.
This thermometer is satisfactory as a defining instrument of the ITS-90 in accordance with the criteria that
W(302.9146 K) 2s 1.11807 or W(234.3156 K) < 0.844235.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 456123
Chief, Chemical Process Metrology Division 1 January 1990
Center for Chemical Technology P.O. No. 123-456
1 January 1990 ITS-90 Table for PRT S/N 1812284 at mA
T(kelvin)
2020202020202020.720.820.921.021.121.221.32121212121.821.9222222222222222222.822.92323232323232323.723.82324242424242424.624.724.824.925.0
W(T)
0.004058540.004131800.004206050.004281300.004357570.004434850.004513150.004592480.004672840.004754240.004836680.004920170.005004710.005090310.005176980.005264710.005353510.005443390.005534350.005626400.005719540.005813780.005909110.006005550.006103100.006201760.006301540.006402440.006504460.006607610.006711890.006817310.006923860.007031560.007140410.007250410.007361550.007473860.007587320.007701950.007817740.007934690.008052820.008172130.008292610.008414260.008537100.008661120.008786330.008912730.00904032
dT/dW(T)
13651.44313467.81613288.05413112.05612939.72412770.96312605.68112443.78912285.20012129.83211977.60211828.43211682.24511538.96811398.52811260.85611125.88310993.54510863.77610736.51610611.70210489.27810369.18610251.37110135.77910022.3579911.0559801.8249694.6159589.3829486.0799384.6629285.0889187.3159091.3028997.0118904.4018813.4378724.0808636.2968550.0508465.3088382.0378300.2058219.7828140.7368063.0387986.6597911.5727837.748
T(kelvin)
252525252525252525.825.926.026.126262626262626.826.9272727272727272727.827.928.02828282828282828.828.929.029.1292929292929,
29.829.930.0
W(T)
0.009040320.009169100.009299070.009430250.009562620.009696190.009830960.009966940.010104120.010242510.010382100.010522910.010664930.010808160.010952610.011098260.011245140.011393230.011542540.011693070.011844810.011997780.012151960.012307370.012464000.012621850.012780920.012941220.013102730.013265470.013429440.013594620.013761030.013928660.014097520.014267590.014438890.014611410.014785150.014960120.015136300.015313700.015492320.015672160.015853220.016035490.016218980.016403690.016589610.016776740.01696508
dT/dW(T)
7765.1617693.7847623.5947554.5647486.6717419.8917354.2017289.5797226.0037163.4517101.9047041.3396981.7396923.0836865.3536808.5306752.5966697.5346643.3276589.9576537.4096485.6676434.7156384.5376335.1206286.4486238.5076191.2846144.7656098.9366053.7856009.3005965.4675922.2755879.7115837.7655796.4265755.6815715.5205675.9335636.9105598.4395560.5125523.1195486.2505449.8965414.0485378.6975343.8355309.452
6.3.6 The following is an example of a Report of Calibration for an RIRT
calibrated over the range from 0.50 K to 30 K on the EPT-76.
138
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
REPORT OF CALIBRATION1976 Provisional 0.5 K to 30 K Temperature Scale (EPT-76)
Rhodium-Iron Resistance ThermometerSerial Number B189
Submitted by
NIST - Thermometry GroupGaithersburg, MD 20899
This resistance thermometer was calibrated by comparison under isothermal conditions with a set of stable
reference rhodium-iron resistance thermometers. A table listing thermometer resistance R as a function
of temperature T follows. This table constitutes the calibration.
In addition, as a convenience to the user, least-squares polynomial smoothings of the calibration data have
been done. A table of temperature, resistance, and the first and second derivatives of resistance with
respect to temperature has been generated from a selected polynomial.
This thermometer was calibrated using an AC bridge at a frequency of 30 Hz using excitation currrents
of 0.125 mA below 1 K and 0.25 mA above 1 K. The current was maintained through the thermometer
only during the actual time when measurements of resistance were made.
For the Director,
National Institute of Standards and Technology
B. W. Mangum, Physicist Test No. 12345
Temperature and Pressure Division 29 December 1989
Center for Chemical Technology Purchase Order No. 123456789
* RHODIUH-IRON THERMOMETER B189
* Calibrated for NIST—Therao&etrY Group
* Calibrated on EPT-76, Mar-June 1988
Calibration Current = 0.125, 0.25 aA
Height = (dT/dR)«2
T(K> R(oheis) Weight I(iA)
0.43976 4.293722 2.87 0.125
0.51872 4.340174 2.86 0.125
0.64985 4.417578 2.87 0.125
0.85100 4.536304 2.89 0.125
1.02827 4.640395 2.92 0.25
1.18271 4.730482 2.96 0.25
1.30962 4.804059 3.00 0.25
1.72215 5.039635 3.15 0.25
2.17154 5.288860 3.37 0.25
2.47997 5.454826 3.55 0.25
3.41404 5.930224 4.22 0.25
4.22195 6.308779 4.93 0.25
4.50339 6.433776 5.21 0.25
5.76304 6.953086 6.64 0.25
7.19977 7.475341 8.62 0.25
8.78512 7.980605 11.2 0.25
10.45483 8.447636 14.5 0.25
12.14414 8.865576 18.4 0.25
13.80845 9.233974 22.7 0.25
15.42337 9.557786 27.3 0.25
17.03860 9.854363 32.2 0.25
18.65600 10.128928 37.3 0.25
20.27451 10.385641 42.2 0.25
21.84940 10.621898 46.6 0.25
23.30557 10.831357 49.9 0.25
24.55872 11.006661 52.1 0.25
25.54642 11.142596 53.4 0.25
26.25499 11.239309 54.0 0.25
26.69840 11.299615 54.2 0.25
26.95381 11.334286 54.3 0.25
27.10184 11.354377 54.3 0.25
27.21302 11.369470 54.3 0.25
09/08/88 -ERP
12th order R-T fit, «>.d. = 0.11 bK
05/10/90 - 11:09:50
RIB189 RESIST-841023
RHODIUM-IRON THERMOMETER B189
Calibrated for NIST—Thermometry Group
Calibrated on EPT-76, Hay-June 1988
FORTRAN POLYFIT
AN ORTHONORMAL LEAST-SQUARES FITTING OF THE POLYNOMIAL
J=N J
R = SUM A T, N = P, P+l, P+2 ....
N J=0 J,N
The fitting is done in double-precision arithmetic. The computer
is programed to select the best polynomial fitting from which to
generate a table of temperature^ resistance* and the first and second
derivatives of resistance with respect to temperature. The page
listing the coefficients of the selected polynomial is clearly marked
at the bottom.
The deviations of each fitted polynomial from the calibration
data are given in the columns headed "Delta R" and "Delta T". The
first and second derivatives of each fitted polynomial (in terms of
the fitted variables) may be given also.
This analysis and the resultant generated table are provided for
convenience in interpolating between the calibration points and do not
constitute the calibration.
This is a fitting of EPT-76 calibration data for the range
0.5 K to 27.1 K.
Table derived from lowest order fit where the ratio
of the coefficients to their standard deviations exceeds 2.00
and the absolute value of each delta T is less than 0.25 mK.
Read in 32 data Foints. 32 of which used for fit.
RHODIUH-IRON THERMOTETER B189
Order of fit = 11
Standard
Coefficients Deviation Ratio
AO = 0.403461344715970+01 0. 251940-03 16014.35
A 1 = 0.58409816514135D+0O 0. 642090-03 909.68
A2 = 0.15633522974138D-01 0.56019D-03 27.91
A3 =-0.125241885614870-01 0.23576D-O3 53.12
A4 = 0.22840914978526D-02 0.56075D-O4 40.73
A5 =-0.254865085790450-03 0.822090-05 31.00
A6 = 0.191971332104650-04 0.778380-06 24.66
A7 =-0.991992563931580-06 0.485050-07 20.45
A8 = 0.346403512658310-07 0.197670-08 17.52
A9 =-0.780615777385770-09 0.507010-10 15.40
AlO = 0.102395491343720-10 0.742660-12 13.79
All =-0.593574267764460-13 0.473690-14 12.53
Standard Deviation of Fit = 0.374764E-04 Oh«s
T R R Calculated Delta R Delta T Weight
Kelvins Ohms Onus Onus Kelvins
0.4398 4.29372 4.29352 0.00021 -0.348 2.87000
0.5187 4.34017 4.34021 -0.00004 0.063 2.86000
0.6499 4.41758 4.41773 -0.00016 0.264 2.87000
0.8510 4.53630 4.53638 -0.00007 0.121 2.89000
1.0283 4.64040 4.64042 -0.00002 0.042 2.92000
1.1827 4.73048 4.73051 -0.00003 0.047 2.96000
1.3096 4.80406 4.80407 -0.00001 0.019 3.00000
1.7221 5.03964 5.03960 0.00003 -0.053 3.15000
2.1715 5.28886 5.28877 0.00009 -0.173 3.37000
2.4800 5.45483 5.45471 0.00011 -0.210 3.55000
3.4140 5.93022 5.93032 -0.00009 0.188 4.22000
4.2220 6.30878 6.30881 -0.00003 0.076 4.93000
4.5034 6.43378 6.43378 -0.00001 0.017 5.21000
5.7630 6.95309 6.95310 -0.00001 0.028 6.64000
7.1998 7.47534 7.47532 0.00002 -0.066 8.62000
8.7851 7.98061 7.98056 0.00004 -0.149 11.20000
10.4548 8.44764 8.44767 -0.00004 0.134 14.50000
12.1441 8.86558 8.86561 -0.00003 0.141 18.39999
13.8085 9.23397 9.23394 0.00003 -0.139 22.70000
15.4234 9.55779 9.55776 0.00003 -0.134 27.29999
17.0386 9.85436 9.85438 -0.00002 0.118 32.20000
18.6560 10.12893 10.12895 -0.00002 0.143 37.29999
20.2745 10.38564 10.38562 0.00002 -0.127 42.20000
21.8494 10.62190 10.62188 0.00002 -0.145 46.59999
23.3056 10.83136 10.83139 -0.00003 0.221 49.89999
24.5587 11.00666 11.00666 -0.00000 0.023 52.09999
25.5464 11.14260 11.14257 0.00003 -0.183 53.39999
26.2550 11.23931 11.23931 0.00000 -0.025 54.00000
26.6984 11.29962 11.29963 -0.00001 0.083 54.20000
26.9538 11.33429 11.33431 -0.00002 0.151 54.29999
27.1018 11.35438 11.35438 -0.00001 0.040 54.29999
27.2130 11.36947 11.36945 0.00002 -0.166 54.29999
Standard Deviation of R = 0.00160 Percent
Standard Deviation of T = 0.18001 nKelvins
RHODItfHRON THERMOMETER B189
Order of fit = 12
Standard
Coefficients Deviation Ratio
A = 0.40353440386808D+01 0. 181260-03 22262.26
A 1 = 0.581681 249467860+00 0.51824D-03 1122.41
A 2 = 0.183151225094050-01 0.51786D-03 35.37
A 3 =-0.13936140688461D-Ol 0.25342D-03 54.99
A 4 = 0.27019102278194D-02 0.71194D-O4 37.95
A 5 =-0.33123326008865D-03 0. 125400-04 26.41
A 6 = 0.28300891747533IH)4 0.14562D-05 19.44
A 7 =-0.172026884480610-05 0.11436D-O6 15.04
A 8 = 0.74047308522468D-07 0.61099D-08 12.12
A 9 =-0.22037748481731D-08 0.21880D-09 10.07
A10 = 0.43099007915774D-10 0.50258D-U 8.58
All =-0. 4980341471 1131D-12 0.66912D-13 7.44
A12 = 0.25757636010739D-14 0.39257D-15 6.56
***** THESE COEFFICIENTS SELECTED FOR THE GENERATED TABLE *****
Standard Deviation of Fit = 0.212764E-04 Ohas
T
Kelvins
0.4398
0.5187
0.6499
0.8510
1.0283
1.1827
1.3096
1.7221
2.1715
2.4800
3.4140
4.2220
4.5034
5.7630
7.1998
8.7851
10.4548
12.1441
13.8085
15.4234
17.0386
18.6560
20.2745
21.8494
23.3056
24.5587
25.5464
26.2550
26.6984
26.9538
27.1018
27.2130
R
Ohas
4.29372
4.34017
4.41758
4.53630
4.64040
4.73048
4.80406
5.03964
5.28886
5.45483
5.93022
6.30878
6.43378
6.95309
7.47534
7.98061
8.44764
8.86558
9.23397
9.55779
9.85436
10.12893
10.38564
10.62190
10.83136
11.00666
11.14260
11.23931
11.29962
11.33429
11.35438
11.36947
R Calculated
Otns
4.29360
4.34024
4.41771
4.53631
4.64035
4.73046
4.80404
5.03964
5.28884
5.45480
5.93033
6.30876
6.43372
6.95307
7.47537
7.98059
8.44763
8.86558
9.23397
9.55779
9.85436
10.12893
10.38564
10.62189
10.83136
11.00666
11.14260
11.23931
11.29961
11.33429
11.35438
11.36947
Delta R
Ohas
0.00013
-0.00007
-0.00013
-0.00001
0.00004
0.00002
0.00002
-0.00001
0.00002
0.00003
-0.00011
0.00002
0.00005
0.00001
-0.00003
0.00001
0.00001
-0.00001
0.00000
0.00000
0.00000
-0.00000
-0.00000
0.00001
-0.00000
0.00000
-o.ooooo
-0.00000
0.00001
-0.00000
-0.00000
0.00000
Delta T
Kelvins
-0.212
0.112
0.215
0.009
-0.070
-0.037
-0.034
0.010
-0.028
-0.053
0.217
-0.035
-0.119
-0.029
0.096
-0.043
-O.019
0.025
-0.001
-0.003
-0.008
0.003
0.019
-0.035
0.034
-0.023
0.006
0.020
-0.049
0.029
0.025
-0.023
Height
2.87000
2.86000
2.87000
2.89000
2.92000
2.96000
3.00000
3.15000
3.37000
3.55000
4.22000
4.93000
5.21000
6.64000
8.62000
11.20000
14.50000
18.39999
22.70000
27.29999
32.20000
37.29999
42.20000
46.59999
49.89999
52.09999
53.39999
54.00000
54.20000
54.29999
54.29999
54.29999
Standard Deviation of R = 0.00114 Percent
Standard Deviation of T = 0.10222 Kelvins
RHODIUM-IRON THERMOMETER B189
12th order
First Second
Temperature Resistance Derivative Derivative
Kelvins Ohms Ohnts/K Ohns/K/K
0.4398 4.29360 0.5906 0.0056
0.5187 4.34024 0.5908 0.0011
0.6498 4.41770 0.5905 -0.0057
0.8510 4.53631 0.5884 -0.0147
1.0283 4.64035 0.5852 -0.0214
1.1827 4.73046 0.5815 -0.0264
1.3096 4.80404 0.5779 -0.0299
1.7221 5.03964 0.5637 -0.0386
2.1715 5.28884 0.5450 -0.0441
2.4800 5.45480 0.5310 -0.0462
3.4140 5.93033 0.4871 -0.0469
4.2219 6.30876 0.4502 -0.0441
4.5034 6.43372 0.4379 -0.0428
5.7630 6.95307 0.3880 -0.0364
7.1998 7.47537 0.3406 -0.0298
8.7851 7.98059 0.2983 -0.0239
10.4548 8.44763 0.2625 -0.0192
12. 1441 8.86558 0.2334 -0.0155
13.8084 9.23397 0.2101 -0.0126
15.4234 9.55779 0.1915 -0.0104
17.0386 9.85436 0.1762 -0.0086
18.6560 10.12893 0.1638 -0.0069
20.2745 10.38564 0.1539 -0.0054
21.8494 10.62189 0.1465 -0.0040
23.3056 10.83136 0.1415 -0.0029
24.5587 11.00666 0.1385 -0.0019
25.5464 11.14260 0.1369 -0.0013
26.2550 11.23931 0.1361 -0.0008
26.6984 11.29961 0.1358 -0.0005
26.9538 11.33429 0.1357 -0.0003
27.1018 11.35438 0.1357 -0.0002
27.2130 11.36947 0.1357 -0.0001
RHODIUM-IRON THERMOMETER B189
12th order
First Second
Tewerature Resistance Derivative Derivative
Kelvins Ohns Ohms/K Ohms/K/K
0.5000 4.32918 0.5908 0.0022
1.0000 4.62380 0.5858 -0.0204
1.5000 4.91350 0.5718 -0.0344
2.0000 5.19472 0.5524 -0.0424
2.5000 5.46542 0.5301 -0.0462
3.0000 5.72463 0.5066 -0.0474
3.5000 5.97202 0.4830 -0.0467
4.0000 6.20776 0.4600 -0.0451
4.5000 6.43224 0.4381 -0.0428
5.0000 6.64602 0.4173 -0.0403
5.5000 6.84973 0.3978 -0.0377
6.0000 7.04401 0.3795 -0.0352
6.5000 7.22948 0.3625 -0.0328
7.0000 7.40672 0.3467 -0.0306
7.5000 7.57632 0.3319 -0.0285
8.0000 7.73877 0.3181 -0.0266
8.5000 7.89456 0.3052 -0.0249
9.0000 8.04414 0.2932 -0.0232
9.5000 8.18790 0.2820 -0.0217
10.0000 8.32623 0.2715 -0.0203
10.5000 8.45947 0.2616 -0.0190
11.0000 8.58794 0.2524 -0.0179
11.5000 8.71196 0.2437 -0.0168
12.0000 8.83178 0.2356 -0.0157
12.5000 8.94767 0.2280 -0.0148
13.0000 9.05985 0.2208 -0.0139
13.5000 9.16856 0.2141 -0.0131
14.0000 9.27399 0.2077 -0.0124
14.5000 9.37632 0.2017 -0.0116
15.0000 9.47575 0.1961 -0.0110
15.5000 9.57243 0.1907 -0.0103
16.0000 9.66652 0.1857 -0.0097
16.5000 9.75818 0.1810 -0.0092
17.0000 9.84755 0.1765 -0.0086
17.5000 9.93477 0.1724 -0.0081
18.0000 10.01998 0.1685 -0.0075
18.5000 10.10330 0.1648 -0.0070
19.0000 10.18486 0.1614 -0.0065
19.5000 10.26479 0.1583 -0.0061
20.0000 10.34320 0.1554 -0.0056
20.5000 10.42021 0.1527 -0.0052
21.0000 10.49592 0.1502 -0.0047
21.5000 10.57046 0.1480 -0.0043
22.0000 10.64391 0.1459 -0.0039
22.5000 10.71639 0.1440 -0.0035
23.0000 10.78799 0.1424 -0.0031
23.5000 10.85882 0.1409 -0.0027
24.0000 10.92896 0.1397 -0.0023
24.5000 10.99852 0.1386 -0.0020
25.0000 11.06759 0.1377 -0.0016
RHODIUM-IRON THERMOMETER B189
12th order
First Second
Te»perature Resistance Derivative Derivative
Kelvins Onus Ohms/K Ohns/K/K
25.5000 11.13624 0.1370 -0.0013
26.0000 11.20457 0.1364 -0.0010
26.5000 11.27264 0.1360 -0.0007
27.0000 11.34056 0.1357 -0.0003
27.5000 11.40841 0.1357 0.0003
6.3.7 The following is an example of a Report of Calibration for an RIRT (theRIRT of sec. 6.3.6) calibrated over the range from 0.65 K to 25 K on the ITS -90.
147
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MARYLAND 20899
REPORT OF CALIBRATIONInternational Temperature Scale of 1990 (rTS-90)
Rhodium-Iron Resistance Thermometer
Serial Number B189
Submitted by
NIST - Thermometry GroupGaithersburg, MD 20899
This resistance thermometer was calibrated by comparison under isothermal conditions with a set of stable
reference rhodium-iron resistance thermometers. A table listing thermometer resistance R as a function
of temperature T follows. This table constitutes the calibration.
In addition, as a convenience to the user, least-squares polynomial smoothings of the calibration data have
been done. The values of the various polynomial coefficients, their standard deviations, and the deviation
of the calibration values from the polynomial values are reported. Finally, a table of temperature,
resistance, and the first and second derivatives of resistance with respect to temperature has been generated
from a selected recommended polynomial.
This thermometer was calibrated using an AC bridge at a frequency of 30 Hz using excitation currrents
of 0.125 mA below 1 K and 0.25 mA above 1 K. The current was maintained through the thermometer
only during the actual time when measurements of its resistance were made.
For the Director,
National Institute of Standards and Technology
Dr. Hratch G. Semerjian Test No. 12345
Chief, Chemical Process Metrology Division 7 May 1990
Center for Chemical Technology Purchase Order No. 123456789
* RHODIUM-IRON THERMOMETER B189
* Calibrated for NIST—ThernoaetrY Group
* Calib. on EPT-76, converted to ITS-90
Calibration Current = 0.125. 0.25 oA
Height = <dT/dR)**2
T(K) R(ohns) Height KaA)
0.51872 4.340174 2.86 0.125
0.64985 4.417578 2.87 0.125
0.85100 4.536304 2.89 0.125
1.02826 4.640395 2.92 0.25
1.18270 4.730482 2.96 0.25
1.30961 4.804059 3.00 0.25
1.72213 5.039635 3.15 0.25
2.17151 5.288860 3.37 0.25
2.47994 5.454826 3.55 0.25
3.41397 5.930224 4.22 0.25
4.22185 6.308779 4.93 0.25
4.50328 6.433776 5.21 0.25
5.76285 6.953086 6.64 0.25
7.19948 7.475341 8.62 0.25
8.78469 7.980605 11.2 0.25
10.45422 8.447636 14.5 0.25
12.14331 8.865576 18.4 0.25
13.80738 9.233974 22.7 0.25
15.42204 9.557786 27.3 0.25
17.03697 9.854363 32.2 0.25
18.65405 10.128928 37.3 0.25
20.27221 10.385641 42.2 0.25
21.84673 10.621898 46.6 0.25
23.30253 10.831357 49.9 0.25
24.55534 11.006661 52.1 0.25
25.54277 11.142596 53.4 0.25
05/07/90 - 14:03:47
EXAMPL RESIST-841023
RHODIUM-IRON THERMOMETER B189
Calibrated for NIST—Thermoaetry Group
Calib. on EPT-76, converted to ITS-90
FORTRAN POLYFIT
ORTHONORNAL LEAST-SQUARES FITTING OF THE POLYNOMIAL
J=N J
R = SUM A T, N = P, P+l, P+2
N J=0 J,N
The fitting is done in double-precision arithmetic. The computer
is programed to select the best polynomial fitting froi which to
generate a table of temperature, resistance, and the first and second
derivatives of resistance with respect to temperature. The pa9e
listing the coefficients of the selected polynomial is clearly narked
at the bottom.
The deviations of each fitted polynomial from the calibration
data are given in the columns headed "Delta R" and "Delta T". The
first and second derivatives of each fitted polynomial (in terns of
the fitted variables) my be given also.
This analysis and the resultant generated table are provided for
convenience in interpolating between the calibration points and do not
constitute the calibration.
Special coments:
(1) This example includes the fitting results of only the
best polynomial fitting. Normal 1y> fitting results for several
lower-order polynomial fittings are included for comparison.
(2) The data are weighted by a factor <dT/dR)«2.
Table derived from lowest order fit where the ratio
of the coefficients to their standard deviations exceeds 2.00
and the absolute value of each delta T is less than 0.20 aK.
Read in 26 data points, 26 of which used for fit.
RHODIUM-IRON THERMOMETER B189
11th order
First Second
Tewerature Resistance Derivative Derivative
Kelvins 0hn5 Ohi»s/K Otas/K/K
0.5000 4.32904 0.5915 0.0004
1.0000 4.62383 0.5859 -0.0210
1.5000 4.91353 0.5718 -0.0345
2.0000 5.19472 0.5524 -0.0423
2.5000 5.46541 0.5301 -0.0462
3.0000 5.72463 0.5067 -0.0473
3.5000 5.97205 0.4831 -0.0467
4.0000 6.20781 0.4601 -0.0451
4.5000 6.43231 0.4381 -0.0428
5.0000 6.64610 0.4173 -0.0403
5.5000 6.84981 0.3978 -0.0378
6.0000 7.04409 0.3795 -0.0352
6.5000 7.22955 0.3625 -0.0328
7.0000 7.40680 0.3467 -0.0306
7.5000 7.57640 0.3319 -0.0285
8.0000 7.73887 0.3181 -0.0266
8.5000 7.89468 0.3053 -0.0249
9.0000 8.04427 0.2932 -0.0232
9.5000 8.18805 0.2820 -0.0217
10.0000 8.32639 0.2715 -0.0203
10.5000 8.45964 0.2616 -0.0191
11.0000 8.58813 0.2524 -0.0179
11.5000 8.71215 0.2438 -0.0168
12.0000 8.83197 0.2356 -0.0157
12.5000 8.94786 0.2280 -0.0148
13.0000 9.06005 0.2208 -0.0139
13.5000 9.16877 0.2141 -0.0131
14.0000 9.27421 0.2077 -0.0123
14.5000 9.37656 0.2017 -0.0116
15.0000 9.47600 0.1961 -0.0110
15.5000 9.57269 0.1908 -0.0103
16.0000 9.66680 0.1857 -0.0097
16.5000 9.75847 0.1810 -0.0092
17.0000 9.84785 0.1766 -0.0086
17.5000 9.93507 0.1724 -0.0081
18.0000 10.02028 0.1685 -0.0075
18.5000 10.10361 0.1649 -0.0070
19.0000 10.18518 0.1615 -0.0065
19.5000 10.26511 0.1583 -0.0061
20.0000 10.34354 0.1554 -0.0056
20.5000 10.42056 0.1527 -0.0052
21.0000 10.49630 0.1502 -0.0047
21.5000 10.57085 0.1480 -0.0043
22.0000 10.64431 0. 1459 -0.0039
22.5000 10.71680 0.1441 -0.0035
23.0000 10.78841 0.1424 -0.0031
23.5000 10.85924 0.1410 -0.0027
24.0000 10.92940 0.1397 -0.0023
24.5000 10.99899 0.1387 -0.0019
25.0000 11.06809 0.1377 -0.0018
RHODIUM-IRON THERMOMETER B189
11th order
First Second
Tenperature Resistance Derivative Derivative
Kelvins Ohns Ohns/K Ohms/K/K
0.5187 4.34011 0.5915 -0.0006
0.6498 4.41765 0.5910 -0.0070
0.8510 4.53631 0.5887 -0.0156
1.0283 4.64038 0.5853 -0.0220
1.1827 4.73049 0.5815 -0.0267
1.3096 4.80407 0.5779 -0.0301
1.7221 5.03965 0.5636 -0.0386
2.1715 5.28882 0.5449 -0.0440
2.4799 5.45477 0.5310 -0.0461
3.4140 5.93032 0.4871 -0.0469
4.2218 6.30878 0.4502 -0.0441
4.5033 6.43374 0.4380 -0.0428
5.7628 6.95308 0.3880 -0.0364
7.1995 7.47536 0.3407 -0.0297
8.7847 7.98059 0.2983 -0.0239
10.4542 8.44765 0.2625 -0.0192
12.1433 8.86558 0.2334 -0.0155
13.8074 9.23396 0.2101 -0.0126
15.4220 9.55779 0.1916 -0.0104
17.0370 9.85437 0.1762 -0.0086
18.6540 10.12892 0.1638 -0.0069
20.2722 10.38564 0.1539 -0.0054
21.8467 10.62190 0.1465 -0.0040
23.3025 10.83135 0.1415 -0.0028
24.5553 11.00666 0.1386 -0.0019
25.5428 11.14260 0.1368 -0.0020
RHODIUM-IRON THERMOMETER 0189
Order of fit = 11
Standard
Coefficients , Deviation Ratio
A = Q.40345618512987D+01 0. 172030-03 23452.05
A 1 = 0.58376417022443D+00 0.42964D-03 1358.72
A 2 = 0. 16272867175907D-01 0.38192D-03 42.61
A 3 =-0.129180383967580-01 0.16723D-03 77.25
A 4 = 0.24053622765144D-02 0.41881D-04 57.43
A 5 =-0.276531060638970-03 0.65102D-05 42.48
A 6 - 0. 216099359574820-04 0.65633D-06 32.93
A 7 =-0.116478740572690-05 0.43662D-07 26.68
A 8 = 0.42603165622150D-07 0. 190260-08 22.39
A 9 =-0.100883806745190-08 0.522350-10 19.31
A10 = 0.139413592145660-10 0.81948D-S2 17.01
All =-0.853183872150180-13 0.56000D-14 15.24
Standard Deviation of Fit = 0.198756E-04 Ohis
T R R Calculated Delta R Delta T Height
Kelvins Ohos Otns Ohms etKelvins
0.5187 4.34017 4.34011 0.00006 -0.105 2.86000
0.6499 4.41758 4.41765 -0.00007 0.116 2.87000
0.8510 4.53630 4.53631 -0.00001 0.018 2.89000
1.0283 4.64040 4.64038 0.00002 -0.026 2.92000
1.1827 4.73048 4.73049 -0.00001 0.019 2.96000
1.3096 4.80406 4.80407 -0.00001 0.021 3.00000
1.7221 5.03964 5.03965 -0.00001 0.019 3.15000
2.1715 5.28886 5.28882 0.00004 -0.069 3.37000
2.4799 5.45483 5.45477 0.00006 -0.104 3.55000
3.4140 5.93022 5.93032 -0.00009 0.187 4.22000
4.2219 6.30878 6.30878 0.00000 -0.005 4.93000
4.5033 6.43378 6.43374 0.00003 -0.074 5.21000
5.7629 6.95309 6.95308 0.00001 -0.017 6.64000
7.1995 7.47534 7.47536 -0.00001 0.044 8.62000
8.7847 7.98061 7.98059 0.00002 -0.054 11.20000
10.4542 8.44764 8.44765 -0.00001 0.038 14.50000
12.1433 8.86558 8.86558 -0.00000 0.020 18.39999
13.8074 9.23397 9.23396 0.00001 -0.055 22.70000
15.4220 9.55779 9.55779 -0.00000 0,015 27.29999
17.0370 9.85436 9.85437 -0.00001 0.041 32.20000
18.6541 10.12893 10.12892 0.00001 -0.035 37.29999
20.2722 10.38564 10.38564 0.00000 -0.011 42.20000
21.8467 10.62190 10.62190 -0.00001 0.039 46.59999
23.3025 10.83136 10.83135 0.00000 -0.033 49.89999
24.5553 11.00666 11.00666 -0.00000 0.015 52.09999
25.5428 11.14260 11.14260 0.00000 -0.003 53.39999
Standard Deviation of R = 0.00081 Percent
Standard Deviation of T = 0.08401 aKelvins
*«** THESE COEFFICIENTS SELECTED FOR THE (BERATED TABLE *****
6.3.8 The following is an example of a Report of Calibration for a thermocouplecalibrated over the range from °C to 1450 °C.
154
UNITED STATES DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
GAITHERSBURG, MARYLAND
REPORT OF CALIBRATIONTYPE S THERMOCOUPLE
(Tagged: S/N 2211998)
Submitted by
ZZYYXX CompanyHometown, U.S.A
All temperatures in this report are given in degrees Celsius (ITS-90). The International Temperature Scale of 1990
(ITS-90) was adopted by the International Committee of Weights and Measures at its meeting in September 1989,
and is described in The International Temperature Scale of 1990", Metrologia 27, No. 1, 3-10 (1990).
The thermocouple was calibrated by comparison with a standard type S thermocouple in the range °C to 1100 °C.
Values above 1100 °C were obtained by extrapolation. The calibration procedure is described in Section 3.1 of the
National Institute of Standards and Technology Special Publication 250-35, entitled "NIST Measurement Services:
The Calibration of Thermocouples and Thermocouple Materials". The thermocouple was electrically annealed in
air before testing.
Table 1 gives corresponding values of the emf of the thermocouple in millivolts and the temperature of its
measuring junction in degrees Celsius when the reference junctions are at °C. The uncertainties in the emf values
given in the table are estimated not to exceed the equivalent of 0.5 °C in the range °C to 600 "C and the
equivalent of 0.7 °C in the range 600 °C to 1100 °C; the uncertainties in the emf values increase above 1100 CCto not more than the equivalent of 2 °C at 1450 °C. These uncertainties are discussed in Section 5.2 of
NIST SP 250-35. Table 2 gives the coefficients of the equations that were used to compute the emf values given
in table 1. Each equation is valid only within the specified temperature range.
The calibration of a thermocouple may change during use. The magnitude of the change depends upon such factors
as the temperature, the length of time, and the conditions under which it is used. Factors affecting the performance
of platinum-rhodium versus platinum thermocouples are discussed in NIST SP 250-35 and in "Accurate
Thermocouple Thermometry", High Temperatures - High Pressures 11, 173-192 (1979). Copies of these documents
are available from NIST, Chemical Process Metrology Division, Gaithersburg, MD 20899.
For the Director
National Institute of Standards and Technology
Dr. Hratch G. Semerjian
Chief, Chemical Process Metrology Division
Center for Chemical Technology
National Measurement Laboratory
P.O. No. 123ABC123
Test No. 998877
Date : May 4, 1990
ooCOo\ON
oPid
O 1^- VO CO CMin o vo cm onOHHNN
0\o o o o
r- in * in r^
co ^ «* in vo
H r- co h orH CO VO ^ CMh h co m o
o © o o o O O O O H
OIN't > Ho co vo ^t coH H CM co >*
<-4 <-i r-4 t-4 t-4
VO CM ON VO •>*
HOCO MOin vo vo t-- co
• • • • •
H H H H rH
CO CM CM CO tm ^ co cm hmoHNnH (M CM CM CM
OOo
0)
X!pCO
a)
R
d)
J3P
a)
M
Prd
U0>
BQ)
P
-p01
EhCOHS3
(0
co•H4JO
0>
ocCD
M0)
0)
pO>
•He
c•rH
<M
0)
CO
vo
in
I'riONinit O VO CN COOHHNN• • • • t
o o o o o
co in <* vo onnoiioHh© O H CM CM
• • • • •
o o o o o
co on co on cmco co ^* o r^OOHMMo o o o o
IN CO >* O VOO O H CM CM
<W
o o o o o
cm co vo r-» ocm r* en on voO O rH H CM
o o o o o
i-i oo r~ co oin h co in nn Tf rf in vo
• • • • •
o o o o o
»* H O H r>>* H co in cmco ^ ** in vo
• I • •
o o o o o
co in co ^ voco o r^ <* Hm rr ^* in vo
o o o o o
h co vo r^ conononon n ^ in vo
^ on vo co cmo t-» in co Hr» r» co on oo o o o H
vo <m co in ^fon r- <* cm ovo r» co on o• • • • •
O O O © H
en ^r o co voCO VO •<*• H ONvo r* co on on
o o © o o
CM VO CM O COco in co H covo r- co on <r>
cm co vo on coon t^ in r» cmo H cm co -*
H fH rH H H
't in h o 'tco vo ^t r> HO rH CM CO t*
H rH rH H H
VO t-» ON CM VOMnnrjoO H CM CO <*•
co n o >ino on co vo inin in vo r^ co
<* n cm H oON O H CM CO
H t-l i-l H H
on in h co voon co t-* in <***• in vo t-» co
H CM CM CM CM
mt't in voCO CM H O CT>
ON O H CM CM
H rH H H H
h vo n o coon r- vo in co^ in vo t^ co
H CM CM CM CM
vo in in in voN H O 01 COC\ O H H CM
co on rH •"t r^VO ** CO rH ONo H cm cn n
H rH rH H fH
cm co >* h onco vo in ** cm't in vo t^ co
rH CM CM CM CM
c- vo vo vo r~HOmcOhfllOOHM
o o o o o
^ rlO>0> HCM on in CM o
o o o o o
<* o\ in CM Hh >* N OCOvo r~ co on on
o o o o o o o o o o
rH H H rH
OrininoiVO ^t CM O COo H cm co co
• « • • •
H H rH H H
H H H H rH
*tf ON in CM Or^ in "3" co cm>* in vo r-> co
H CM CM CM CM
co f» t-» t*» ooo on co r^ voon on o h cm
rH H CM CM CM
0! o> cr>
<rl COn EhQ) H<w
0)
3(0 •H0) W
HH0)
A3 U>
V)
0)• a)
H u0>
30)
mrt! eEH •H
o 0)
o\ r<
a\ 3rH p
fij
«. u't 0)
>i 6n) 0>
S Eh
CM
u
vo cm o H nH r^ r> en inO O rH H CM
o o o o o
h vo'jinhH VO CM CO «tf
O O H H CM• • I • •
o o o © ©
mnoxoH0\OHh<fO O H H CM
• • • I •
o © o o ©
o in r> cm •*
oinHhn© © H H CM
C • * • •
o o o o o
© © © © ©H CM n "*
CO <tf CM CM "tf
h co m CM on n 'j- in in
© © © o ©
H co vo in Is"Hhf HCO
r> n >* in in
© o o o o
in h o\ co eno t^ n o r^nnf mm© o © © o
00 <tf CM H CMo> vo r> © r»cm n t in in
a e c « •
© o © o o
© o o o ©in vo r^ co on
t*» h r» "* nvo ^ h en t-vo r» oo co on
© © © o ©
o^or^mV0 CO rH CO VOvo (>• CO CO ON
o © © o ©
CM VO CM ON t-»
in NOhinVO (^ CO CO ON
© © O © O
in ON "* rH ON=* h on r- ^vo r- r^ co on
© o o o o
o © o © oOrl(Mn<*rH H H H H
CMn<*>Hin n H on coo H cm cm nrlHHrlrl
t? in vo on cm<t rjocohO H CM cm n^ r-i r-i r-i ^
VO [-» CO o •<*
n H ON CO voO rH H CM CO
rH rH rH rH rH
00 CO © CM incm © o\ r^ in© rH H CM nrH H rH rH rH
o © o o oin vo r» oo onrH f-i rH H H
in o r^ cn Hvo in n cm h"# in vo r^ co
H H rH H H
r-^ cm co in CMin ^ cm h o-a* in vo r- co
H rH H H H
co r> on vo <*
^ n h o on<* in vo r^ r^
rH t-i H H rH
OlOHMn•t CM H ON CO* in vo vo hrH H r-i H H
© © O O OO H CM CO "*CM CM CM CM CM
ON 00 00 00 ONon co r^ vo inCO ON O H CM
H H CM CM CM
H ON ON ON Oon r- vo in inCO ON O H CM
H rH CM CM CM
CM H O O Hco r^ vo in <*
CO ON O H CM
H H CM CM CM
n CM H H CMt-» vo in ^f r>CO ON O H CM
H H CM CM CM
O O O O Oin vo r» co onCM CM CM CM CM
9OOo
o»
xipwa)
R
(V
XIp
o
a>
5(0
u
&0)
pw
(0
r»ooooencn
Os•p(0
a)
Eh
W
U
•P«J
(0
CO•H-POc
•r-»
at
oca)
a>
<wa)
«
CO
•pHo>-H
•H
C•H
t-l
n0)
<u o> cn
<H CO6 Eh<u H<w
CO
3(0 H0) W3 HH 0)
«J CJ>
t 0)
H u
a0)
CQrtj cEh •rH
O 0)
cn M<T> 3H p
id«. u
<* (0
>i 6«J 01
X EH
en
CO
vo
in
co
CM
u
o cn cn co r--
H VO H VO Ht- vo vo in inoomoHN
t*» <* H oo in"t "t "t ci nr> rr m vo r- comoHN
vovoininvo htomnnN O) CM N (M CM CM CM CO coco ** in vo r-- mwohm
N N N « N
vo co h ^r coen oo co r- von^invoh
NNnnn
n vo h vo Nvo in in <# <*CO Cn O H CM
co co co co co
oo "# h oo inco co co cm cmro Tt tn vo r-
n n<t 't <t
n h en co r^r\i CM H H Hoo cn o h cm
'tf ^ in in in
vovoininvo r-» r-- cn o cnH H r-\ H H HHHCNrjcoTtinvor^ co cn o h cm
NNNNN NNrinn
r> cn cm in coco r> r- vo inn ^ in vo r-»
cm r-» cm r-- cmin *t t n roCO Cn O rH CM
co co co co co
co in h oo voMMNHHn <* in vo >
nri't't <f
H H O O OCO Cn O H CM
* * in in in
vovoininvo vo r- cn o cnOOOOO OOOHHco *r in vo r- co cn o h cm
CM CM CM CM CM CM CM CO CO CO n co co co co
co o cm vo cnr» r-» vo in ><*
co «* in vo t-»
co r~- cm r^ co cnincNcnvo* co co cm cm h h i-h o oco cn o h cm co t in vo r-
nn <f <f ^
(iHOCOho o o cn onco en o o H
vo vo in in voen en en en enrg co t in vo
^j* 't in in in
vo r- en o rgen en en o or*» co en H cn
cm cm cm cm cm cm cm co co co
oiHrnoovo vo in *t ^co **• in vo r«
*i* CO co oo coCO CM CM H Hco en o H cm
CO co co co co
en in cm en voo o o en enco >* in in vo
CO CO ^f ^ ^
fCJOCOhen en en oo coMOO\OH
vo vo in in voCO CO CO CO 00cm co Tf in vo
<* <* <* in in
VO h CO O Nco co co en enr^ oo en o h
CMCMCMCMCM CMCMCOCOCO CO CO CO CO CO CO CO CO *3* *3* ** 'j* •<* in in
e0)
oVOCO
CMin •>*
mCOVO
OCOr-
CM00
COHen
COHO
COo o
CM
enenCM
voenn
CM
rT
01COin
voCOvo
COCMCOCO
OCOcn
CO
or-H
vor-CM
VO
ro
in in
tn
VO
VO
VO
CO
00
cn
O CMCO COO H
CM CM CM CM CM CM CM co CO CO CO CO ro co CO CO ro ro «* •* "t rf >* * -* •>t ^ >* in in
OinCO
CM inCOin
COCMVO
HCM
inHCO
enoenOO
vTi
eno
enOenCM
VO00CO
COCO
VTi
inr-VO
C4
CO
or-cn
cnvoo
r-voH
vovoCM
VOVOro
inVO
invoin
VOVOVO
voVO
Is-
VOCO
COVOcn
O HO H
CM CM CM CM CM CM CM co co CO CO ro ro fO CO CO ro ro x* * X* •tf «* ^ >t * >* •<t in in
H co in co cm** CO CM H Hco ^r in vo r-
in cn rr cn ^to cn cn co coCO CO Cn O rH
o vo co o r»co r> r-« r> vocm co tj< m vo
^ N O 0\ >vo vo vo in inr» co cn o H
vo vo in in vo
in in in in incm co * in vo
vo > co cn hin in in in vor> co cn o H
CM CM CM CM CM
cm *t vo cn cmCO CM H O Oco •** in vo r-
CM CM CM CO CO
vo o in o incn cn co co r*-
r- co cn o H
co ro co co co
h r- co o r^r-» vo vo vo incm co ^ in vo
CO CO CO ^* >*
in cm o cn coin in in -* *r- co cn o h
r- vo in in vo
cm co -^ in vo
^ t t in in
vo r> co cn hTt * ^ ^ inr-> co cn o h
CM CM CM CM CM
co in r- o cocm h o o cnn Tf in vo vo
CM CM CM CO ro
r- h in o inco co r- r^ vor-- co cn o H
co ro ro ro ro
H r- t? o r-vo in in in ^cm co »t in vo
ro co ro "* "*
in co H cn co<5j< Tt Tt ro ror- co cn o h
^f ^j* ^ ^ ^*
r- vo in in inCO CO CO CO COcm co ^ in vo
>* * * in in
vo r- co cn hCO CO CO CO **
r- oo cn o hCM CM CM CM CM
OOOOOO H CM CO "*CO CO CO CO CO
CM CM CM CO CO
OOOOOin vo r- co cnCO CO CO CO CO
ro ro ro co ro
OOOOOOHNdf^T *s7* ^T ^5^ ^J^
ro ro ro ^t ^J*
OOOOOin vo r» co cntj* ^j* ^f ^* ^
OOOOOOrlNClfin in in in in
rr ^t "t in in
OOOOOin vo r> co onin in in in in
r» ar- e
COCO ocncn +J
(0
in
55 c
P •H<D in •PH OJ
ft H c3 aH •HI
coH ai
Bz
cCD OJ
X! MP OJ
CO OJ
OJ
sPOJ in
X! P•P Hh >O •H<w rH
HQ) •HM 63-P C(0 •HMd) <Ha 66 OJ
OJ
-p
w3in *
u .—
.
oj o> CTi
<H CO
s HOJ H<Mo in
2in •H0) in
3 HH OJ
(0 o>
a
in
OJ
OJ
u
OJ
CO.
< cH •H
o OJ
OS Mas 3H -P
ft?
V M•<* OJ
>1 6Rj OJ
s Eh
Cn
CO
VO
in
r>
CM
in co h o r»co co *j* ^ ^*
o ^ co ro com in in vo voCOO\OH«
n cn in h cor~~ t~ co cn asro Tt m vo r-»
in ro H en coOHNNriWOHNrnr» r- r» r- co** in vo r- cotj- in <o r» co
cn h cm •>* r-»
as H cm ro «*
O^ H in n *in in in in in
in r~ o co vocm cm ro co coro «<* in vo r>
in in vo vo vo
o »r co ro r*«tf »* <* in inco en o H IN
vo vo vo vo vo
no* Hhvo vo r> co con •<* in vo r»
vo t-» r^ r- r-»
in rvi o as r-*
cn o h h cmco o H in ro
r- r-» r-» r» r-
r- vo vo vo (-
ro t in vo r^<t in vo r-~ co
r-» co co co co
co o h co inco o h cm ro
in in in in in
H H CM CM CMn^ in vo h
in in vo vo vo
ontONhco ro ro *r •«*
co as o H cm
vo vo vo vo vo
cm co ^ o r-»
in in vo r-» r*n t in vo t^
vr> r- r~- r~ r-
t H 5i » VOCO en Cn O Hco as o cm ro
r» r» r- Is- r-
vo in in in von n t in vorr in vo r- co
r-» co co co co
MO O N^ft^ CO O H CMas o cm ro *
in in in in in
^ r^ o ro voo o H H Hro >* in vo r-
ininvovovo vovovovovo
as ro r- cm voh cm cm ro roco en o H CM
cm r^ ro cn voti< ^r in in von rfin voh
vo vo r^ r^ r^
r> h en r-* vor-» co co cn oco cn o rH ro
r^ r- r^ r- r-
in ff t inHN rrtinTt in vo t- co
r~ co co co co
vo r» en h nvo r^ co o hcn o h ro **•
in in in in in
•^ vo en cm inen en en o ocm co t vo t-<
in in vo vo vo
cn ro r^ h voo rH H <M CMco en o h cm
vo VO vo vo vo
H mn mmnri't't inro *r in vo r»
vo vo r- r- r->
cm o co vo invo r- r^ co enco en o h cm
r» r» r» r- r-
^4* n n n ^i1
o H cm n »*•>* in vo r^ co
r^ co co co co
in vo co o cmin vo r- en offlOHN't
in in in in in in in vo vo vo vo vo vo vo vo vo vo t- r- r- r-» r- r- t^ t^ f** CO CO CO CO
nOJ
COCM
VO
COenCOI-
CMasin
inenvo
COen
CMOen
voooHrHrH
inl-t
CM
HCMn
VOCM
(Mnin
COnVO
in<*
CMmCO
inen
voo
inCOCM
nen
<NOin
CMHVO
CM(N
COCOCO cn
inino
VOVO
HCO Hr^ enCM CO
in in in in in in in VO VO VO vo VO VO 'O vo vo vo r» \> r^ r~ r^ r^ r~- r- t-- CO co CO CO
CO
CM
VO
ro
en CMCOin
inCOVO
COCO
(NenCO
VO
en
ooH
inofM
orHn
inHHCMin
CMVDn
H>*
CO
CO
en
VO
ino
VOH
COr-CN
CMCOCO
rHcn•*
HOVO
H CMCMCO
COCOcn
<*
oininH
r^ envo r>CM CO
in in in in in in in in VO VO vo VO vo VO VO vo vo r^ r> r^ r- r^ r^ r- r^ r» CO CO CO CO
CO VO 00 H xfvo vo vo r^ r^cm co rt in vo
co H in o 'tr^ co co cn cnMOffiOH
cn m H r^ roCn O H rH CMcm ^r in vo r~
o co in co cmro ro ^r in voCO Cn O rH CM
h o o o Ht-» co cn o Hro «* in r^ co
H ro -* vo cocm co <* in vocn o H cm ro
in in in in in
n in co h Ttin in in vo vocm ro o» in vo
in in in vo vo
r^ H m cn *?vo t-« r» t-» cor^ co cn o rH
vo vo vo vo vo
cn •* O vo roco cn o o Hcm ro in vo r*
vo vo r- r- r^
o r^ in ro rHcm cm ro «* inco cn o H CM
r- r» r*» r^ r-
o cn cn cn oVO VD MO Oco »* in vo co
r^ co co co co
o cm ro in r^rH cm co >* incn o h cm ro
ininininin inininvovo
co in co H 't»* •'t ** in incm co ** in vo
r>- h in cn Ttin vo vo vo r-*MOOlOH
VO VO VO VO VO
co *t cn vo cmr- co co cn ocm ro ^r in r^
vo vo > r~ r^
cn vo ^j* cm oOHNn'fco cn O H CM
r» r- r- r> r»
en co co co cn"t in vo r^ con t m vo >
r^ co co co co
cn o cm * vo
cn h cm ro ^CO O H CM ro
in in in in in
o o o o oo H cm ro >>*
VO VO VO VO VO
in in in vo vo
o o o o oin vo r^ co envo vo vo vo vo
vo vo vo vo vo
o o o o oOHNfTfr- f» r^ r» t^
vo vo r^ r~ r^
o o o o oin vo > co cnr- r- r^ r- r-
r» r^ r- r^ r^
o o o o oOHNn^fCO CO CO 00 CO
r~ co co co co
o o o o oin vo r^ co cnCO CO CO CO CO
o mrnooin hHIOHh cm co in H co vo m H cn co vo in ^ n n n cm cm r> r>r» in r» co cn h cm * in c- co o h n in vo co o cn n in t^ cn h n in t^ cn h n in
CO o <nin vo r- co o H m n »f in r- co cn o h cm ^t* in vo h- co cn H cm n * in p* co cn
CO CO CO CO CO T\ cn cn cn CT> cn cn cn cn o o o o o o o O O H H H r4 <-\ <-i rt Hencn
pfd
H H H H H H rH rH H H rH H r-{ <-1 rl r-\ r-i
d(0
COH^fflH vo o in o in h t^ r> o t^ "<* rH cn r^ vo ** n CN H H H O O H H2 ^ VO h- CO O h n t vo t^ cn o cn ^r in r» cn o cm ^t VO 00 O CN tJ- VO 00 O CM ^t
+j-H-P CO
in vo h co o HNO'f in vo co cn o h cm r> in vo r- co cn h cm cn * in r» co cn
(0 CO CO CO CO cn Cf> Cn CTV CTi CT> cn cn cn o o o o o o o O O H H H H H H H Htt a) c H rH i-l i-t H H rH HrlHHH r-i r-i r^ f-i r-f
a H 3o •r->B
CO d) honwo ^ oo n to to cn in H co in NOMOf cm rH o cn cn cn co co cn cnH n in vo r^ en o H m *r vo MJ\H N f vo co cn rH o in r- <n o cm ^r vo CO O CM2 cr>
in vo r» co cn HfMn^ifl vo r~ cn o <-i MPTHOh co cn o cm n ** in vo co cn
CO CO CO CO CO cn cn cn cn cn cn en cn o o o o o o o O O H H rH H H H H Hp a)
«MH rH rH rH rH H H H H rH rH H H H H H H
co
<u vo co cm in cn m t^ cn t^ cn co m o vo n O CO VO ** CM h cn 00 CO f"^ MO lO h hft
cm n in vo r^ worjnin vo co o h n in VO CO O CM t in mji h n in r» cn hvo
in vo t-» co en o cn n rr in vo r- cn o h cm n 't vo r^ co cn o h n rj- in ys r^ cn
^ CO CO CO CO CO en cn en cn en cn cn cn o o o o o o o O O H H H H H H H H<u 05 H H r-\ r4 t-i r-i r-\ r-{ r-{ r^ ,-i r^ H H H H HX! PPU > Tf hOth h vo o in o VO CM co in (M cn vo ^ cm o cn r- vo vo in in tj< ^r rt m
•H h cm *t in vo CO Ol H (M Tt in t-» co o eg ro in t-» cn H cm ^r vo co o CM "* VO CO O<M
rH•H
inin vo t^ co en oHnf in vo r- co o h cm o tt in r^ co cn o h m >* in vo r- cn
(V CO CO CO CO CO Cn Cn ON 0V GV cn cn cn o o o o o o o O O rH rH H r-\ r-\ <-\ rl <-\
Ps rH H rH H rH rH rH rH H rH H H H rH H rH Hc
rd •Hb m vo en cm vo o **• cn ">* en t o vo r> o r- * CM O CO t^- vo in «* n r> cm cm cm na) <M o H <n tt in r^ co en H cm <* vo r> cn H cm f vo co cn h ro in r~ cn Hninhm
(U
s0)
in vo r> co cn O H (N t in vo t^ co cn h cn n t in vo co cn O H CM 't in vo r^ co
B CO CO CO CO CO en en en en en cn cn cn cn o o o o o o O O H H H H H H H Hp01
w
a> H H H H rH H rH rH H H H rH H H H H
^ cn in co h in co r> r> cm t> n cn in h co in n o co vo in >* r> cm H H O O O H>H «-^ woHn^ IOMOOH n ^ vo co cn h r> in vo co O CM <* VO CO O CN "* VO CO
>o0>1
n>tf VO t^ CO CTl OHN-tin vo r» co cn o cm o ^t in vo 00 Cn O rH CM >* in vo r» oo
CO CO CO CO CO en en en en en cn cn cn cn o o o o o o O O H H rH rH H H H H<M CO H r-l rl t-i rl t-< H H H H H r-i rl <-{ <-\ HB EH<U H<u o m vo o m (* i-\ \D r4 KO HhriOiO ^- h cn vo in m cm h o cn cn oo oo co cn
W co cn o cn n *r vo r*» cn o cm n in r^ co o cm n in r» cn h m in vo CO O CM rj- VO
V)
3•Hto
CM* in r> co cn oHnnin vo r~ co cn o cm tn ^ in vo r- cn o h cm o in vo > co
CO CO CO CO CO cn cn cn cn cn cn cn cn cn o o o o o o O O H H H r-i t-{ r-i H r-\
3 rH
o«H H H rH H H H H rH rH rH rl t-\ r4 r4 r-{
fd
>w en cn in co cn vo o >* cn <* o in cm co in cm cn t^ in o h o cn co t^ t^ vo vo vo r»
Q) vo co en o cm r> in vo r- cn h cm -^ in r» cn O CN ^ vo co o rH n in > cn h n in•
Ha)
u H«* in vo co en O rH CN m *3< vo r- co cn o h n ^j* in vo r^ cn o h cm n ^r vo r- oo
CO CO CO CO CO cn cn cn cn cn cn cn en cn o o o o o o O O rH H H H H H H H
a0) H r-^ r-i r-\ t-i f-\ H rH H H H H H H H H
< cEh •H «H^>0 ^ oj n to n cotown o r>- in n H cn co r- vo in in «* >* "* in
in r^ co cn h cm n in vo co cn h n 'S* vo co cn H n in VO CO O CM -* VO 00 O CM TOOS 3
O<* in vo t-» cn O H CM O tJ< in t- co cn o h cm ^ in vo r- 00 O H CM n rt vo r^ co
CO CO CO CO CO cn cn en cn cn cn cn cn cn o o o o o o O O H H rH r-i r-i t-i t-{ r-i
H Pfd
H H rH rH H H t-i H rH H H H H H H H
<* d)
a o o o o o o o o o o o o o o o o o o o o O O O O O O O O O O>i 6 u O H CN O <3" in vo r^ co cn O H cm r» •^, in vo r* co cn O H CM n •* in vo r^ co cnCO 0) e os en en cn cn cn cn cn cn cn o o o o o o o o o o H H H H H H H H H Ha Eh r-H rH H H H rH r-i r-i <-{ rl H H H H H H rH H H rH
ur» O
t>-
CO o00cn +Jen (0
•V)
c55 o
•H<D P PH W oa 0) c53 H 53
o *nHto a>H o
p 55 c0) a>
J3 bPto a)
cn
a
a)
PIH
O<w
a)
P(d
u
&pto
53
Wn
o>•HHiH•HS
C•H
<WBd)
0) o> en
<H COB H0) H<MO to
53
(0 •H0) 0)
53 HH 0)
(0 o>
to
CD• 0>
H rl
IT
90)
CQ
< cH •H
O CD
cn Mcn 53
H P^ u
«* 0)
>i B(0 0)
a H
CO
VO
If)
co
n
O
co <* tn vo r^r>» cn h co inoHn<tin
co cn o cm cor» o> cm •>* voVOMHOH
<* vo r-- en oCO O CM *tf f»n t in vo h
cm co in r^ coen rH co in r»CO O rH CM CO
en H CM CO t*
en cm ^* vo co-* vo r-» co en
cm cm cm cm cmrt r-i r4 t-{ H
cm cm cm co coH H H H H
ro ro co ro cort rl ri r-\ r-i
CO ^* ^ ^ ^*H H H H H
^* ^* ^ ^ ^*<-\ r-i r-{ r-\ r-i
H CM CO ^ inVO CO O CM >*
O H n 1" If)
vo r-» co cn HVO CO O CM If)
*OMJlOHcm >* in t^ COr» en h r> incm ro in vo r-»
onntvo00 O CM ^ vo00 O rH CM CO
t^ cn O H CMco O ro in r^t* vo r^ co en
cm cm cm cm cmt-i r-t t-i r-{ H
cm cm cm n roHHHHH ro ro ro ro roH rH H i-l H
CO 'S* ^ ^f ^1"
rH rH rH rH rH HHHHH
0\OHNN•<*• r« cn h co
•«* in vo r^ cnin r-» cn H r>VO t^ CO O H
o cm n m vovo co o pj "tcm ro in vo r~
co en H CM *tf
vo co h co inco en H cm co
in vo co cn or- (ft Hn vo•<* in r-» co cn
cm cm cm cm cmH t-i H f-t r^
cm cm cm co coHHHHH ro ro ro ro roH H H H H
n n ^ 't "tH H H H H
^J* ^J* ^J* ^y ^yHHHHH
t"» CO cn cn oco m r>. cn cmOHnnin
H co <* in vorf VO CO O CMvo r-» co o h
co en h cm *** vo en rH encm co ^ vo r^
in r- en o cmin t^ en cm "tco en o cm co
CO >* VO t^ 00VO CO O N ^tTt in r» co en
NNNNNH H i-l H H
cm cm cm co coH H i-l H H
(*) f) C"l o oH H H H H
n n ^ ^ ^HHHHH HHHHH
If) vo f» r» COCM ** VO CO oO H CM CO in
mONO'tcm in r^ cn Hvo r-» co en h
vo r^ en o cmr> in r^ o cmcm n ^ vo r^
co in vo co en^ vo CO O CMco en o cn ro
H CM ^r in VOin r- en h co'S* in vo co cn
NNNNNHHHHH CM CM CM CM COH H H i-l rH
n n n o mH rH rH iH i-l
ci n ^ ^ ^HHHHrt r-i <-i t-t t-{ <H
r) <tf ** If) vo
O H CM CO rf
r^ CO O H CMH co vo co oVO h CO (ft H
^t mh co oN 't VO CO Hcm m ^t m r>
Hn -* VOhco in r^ en hco en o H co
cn o h co **CO vo CO O CM»* in vo co cn
B CM CM CM CM CM0) H H H H H
CM CM CM CM COi-l H i-l H H
n o o m nH rH rH H H
CO CO ^1* ^" 'i'HHHHH HHHHH
HNNfTj1
o cm ** vo coo H cm co **
in vo r^ cn oo cm -^ vo envo h> co en o
rH r> t vo r*-Hninh mCM r> <* in vo
en h cm »* inH ^f VO COOco en o h co
r- co cn h cmcm •«* vo cn H^ in vo r- cn
CM CM CM CM CMHHHHH CM CM CM CM COH H H rH rH
n n o n nrH i-H H H H
CO CO 'S' ^ 'tf'HHHHH T^1 rj* ^* ^* ^*HHHHH
OIOOHNco H n in r^
CO ^r If) {-» COcn h co in r*in r^ co cn o
en h cm ^ in01 N t VO COHn<*w vo
r» CO O H COo cm in r- enco en o h cm
't VO> COOh co in r- o* in vo r>- en
H cm cm cm cmH H H H H
CM CM CM CM COH H H H H
O t^t (*) n nHHHHrt CO CO ^" ^* 'i'HHHHH H r-i H t-\ r-t
t^ CO CO CD Or^ cn H co vo
H n n <t voCO O CM <* voin r^ co cn o
r» en o cm nco o n in r^h co •«* in vo
in vo co en hen h co in coMJlOHN
cj t in vo coO CM -* VO CO* in vo r- co
HHrlHH CM CM CM CM nH «-l H i-l H
ro co co co coH H rH H rH
CO CO ^ ^ ^J"HHHHH t-\ r-i r-< t-i t-i
If) If) VO 1^ COVO CO O CM ^moNn<*
en o h cm <*vo en rH r> inin vo co en o
in vo co en hr~ en h co voh cm <* in vo
co ^ vo r^ en00 O CM •tf vor^ en o h cm
o cm co 't inen h co in r~co in vo r^ co
cncn
H CM <N CM CMH H H H H
CM CM CM cm n,-i ,-i ,-i r^ H
CO CO CO CO COH rH rH H rH
CO CO ^f ^f ^r^ <-\ r-i <-{ t-\
^7* ^3^ ^3^ Vr ^J^
H r-{ t-i H i-i H
O O O O OOHNH'tCM CM CM CM CMHHHHH
o o o o oin vo r- co enCM CM CM CM CMH H H H H
O O O O OO rH CM CO •>*
CO CO CO CO COH rH rH H H
O O O O Oin vo r-» co enCO CO CO CO COHHHHH
O O O O OO H CM CO ^t^J* ^3* If ^J* ^^t-\ r-f r-\ r-i <-\
oin<*
H
r» r- •
r- • • ^^H 00
00 Boen
oen
CU <T> •P 1
iH en <H ^-^ CO4a r^ Eh« •
<u cd HEH o
525X!P +
v-' o•
c-P
CO o o n VO CT> VO o o o•H «H VO 3 in o o O O rH o o o
CO O *—
*
•H ** + 1 1 1 1 + + +c 0)
Eho CO H W w w w w www
a) cu cr» t-i cm r^ cm in VO o o o>
EHn P CU o in >* cm vo in o o o
•H «j -—
'
o p cm CTi o r> o o o oo> CO VO in o rsj o rH o o oH
2 cu rd CO CO H <N CO CO o o o oto h CU rH cm r> n o n o o o<D d + CU • t^ rH co r- en o o o3 p u •<* CM in n in in o o orH (0 in 0> VO(0 u ^-» 0) o H n VO H vo o o o> cu o
enTJ H + + + I I + + +
<H g P CB CU ^-* •H0) p in CO
«J a) H H rH r> vo en (N in ena) <H M •• • O O o o o r-l <-< r-\
X! O + 3 cu "* + J 1 I I i i ip P *H VO w w WWW wwwc •<* rd <o o r^ (N Ttf T* >* r> cm H
Q) o ^» M rH o CA co in m CO CO CO•P
a)
H§•
O
H o cu co in in H VO CO o vo r^3 P en ft cu o CO H rH t^ H r^ cm r^
& O P g tr> p r^ ^f r- o in o vo vo
6 c •—'cu c as <N en en in r^ "* cm
O 3 <sr P (0 in tn ^ ^ r> o in vo t^O <M rd M rH t^ r^ in cm ^ VO CO rH
cu VOo •0 + X! CU • r^ V0 cm in vo * H n•p P rH o + 1 + 1 + i + 1
(0 n 3 nT> ft) *—
.
CO P vo<D o -h (d
to <H en Md CO e P ocu o n in co H •<* CO rH
cu >—
»
enft in o o o o H <H H CM0)
&EH
r> p e H + ) 1 1 1 1 1 i
c <H (0 CU VO wwwwwwwwTJ P • o CM CO f"» CJ m co m
•H + c o o n en h H \OOhP (0 rd p n o VO en vo en **• "^ CMid <D CM c vo o in f* P* en r^ h o& 3 ,-^ 10 cu o in co en vo o r> o
H o p rH O o CO co n en in en voa> (0 en rH <D P o CO in in o vo n vo
> •P O <H o n <N r>j CO o co n<H > <H o
CT CM -H -H • o in H fN m cm in HC (0 H TJ o + + 1 + i + +
(0 •H HP P + •H Q) II II II II II II II II
c 3 ex: • •
a> ft ,—» p ^-N Or•H n o c U (d (d td id id to ed <d
o en •H M•H o P O —
-
«H *-* <W <M<H rH H e cu
<U o (0 cu CO tr>
O <H p cCJ
CO
c
+
Ocu cXI cu
p-H(d
• o rd o cu
CM enen
•Hp II
CO -H-H <H 3
3CQ
H 10 <W P•^ &
w W CU (d
<J **• cu CU u cu
H0)
X!EH
rH
cu cu
XI X!£ EH
ncu
Eh
6.3.9 The following is an example of a Report of Calibration for a liquid- in-
glass thermometer calibrated over the range from °C to 100 °C on the
IPTS-68(75).
162
U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
(FORMERLY NATIONAL BUREAU OF STANDARDS)NATIONAL MEASUREMENT LABORATORY
GAITHERSBURG, MD. 2 0899
REPORT OF CALIBRATIONLIQUID-IN-GLASS THERMOMETER
TESTED FOR: NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGYDIVISION 586
MARKED: SURETY TlRANGE: -2 TO +102 DEGREES C IN 0.2 DEGREE
.MOMETER CORRECTIONREADING (IPTS-68) **
-.10 c .10 C20.00 .0540.00 -.0760.00 .0180.00 -.04100.00 .00
**ALL TEMPERATURES IN THIS REPORT ARE BASED ON THE INTERNATIONALPRACTICAL TEMPERATURE SCALE OF 1968, IPTS-68. THIS TEMPERATURE SCALEWAS ADOPTED BY THE INTERNATIONAL COMMITTEE OF WEIGHTS AND MEASURES ATITS MEETING IN OCTOBER, 1968, AND IS DESCRIBED IN "THE INTERNATIONALPRACTICAL TEMPERATURE SCALE OF 1968 AMENDED EDITION OF 1975,"METROLOGIA 12, NO. 1, 7-17 (1976).
ESTIMATED UNCERTAINTIES IN THE ABOVE CORRECTIONS DO NOT EXCEED 0.05DEGREE UP TO 102 DEGREES C .
FOR A DISCUSSION OF ACCURACIES ATTAINABLE WITH SUCH THERMOMETERS SEENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY SP 250-23, LIQUID-IN-GLASS THERMOMETER CALIBRATION SERVICE.
IF NO SIGN IS GIVEN ON THE CORRECTION, THE TRUE TEMPERATURE IS HIGHERTHAN THE INDICATED TEMPERATURE; IF THE SIGN GIVEN IS NEGATIVE, THE TRUETEMPERATURE IS LOWER THAN THE INDICATED TEMPERATURE. TO USE THE CORREC-TIONS PROPERLY, REFERENCE SHOULD BE MADE TO THE NOTES GIVEN BELOW.
THE TABULATED CORRECTIONS APPLY FOR THE CONDITION OF TOTAL IMMERSION OFTHE BULB AND LIQUID COLUMN. IF THE THERMOMETER IS USED AT PARTIAL IMMER-SION, APPLY AN EMERGENT STEM CORRECTION AS EXPLAINED IN THE ACCOMPANYINGSTEM CORRECTION SHEET.
TEST NUMBER 999999P.O. NO. 99999COMPLETED 10-3-89
6.3.10 The following is an example of a Report of Calibration for the liquid-in-glass thermometer of section 6.3.9 calibrated over the range from °C to100 °C on the IPTS-68(75) but converted to the ITS -90.
164
U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
NATIONAL MEASUREMENT LABORATORYGAITHERSBURG, MD. 2 0899
REPORT OF CALIBRATIONLIQUID-IN-GLASS THERMOMETER
TESTED FOR: NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGYDIVISION 586
MARKED: SURETY TlRANGE: -2 TO +102 DEGREES C IN 0.2 DEGREE GRADUATIONS
MOMETER CORRECTIONREADING (ITS-90) **
-.10 C .10 C20.00 .0440.00 -.0860.00 -.0180.00 -.06100.00 -.03
**ALL TEMPERATURES IN THIS REPORT ARE ON THE INTERNATIONAL TEMPERATURESCALE OF 1990, ITS-90. THIS TEMPERATURE SCALE WAS ADOPTED BY THEINTERNATIONAL COMMITTEE OF WEIGHTS AND MEASURES AT ITS MEETING INSEPTEMBER 1989, AND IS DESCRIBED IN ' 'THE INTERNATIONAL TEMPERATURESCALE OF 1990", METROLOGIA 27, NO. 1, 3-10 (1990).
ESTIMATED UNCERTAINTIES IN THE ABOVE CORRECTIONS DO NOT EXCEED 0.05DEGREE UP TO 102 DEGREES C .
FOR A DISCUSSION OF ACCURACIES ATTAINABLE WITH SUCH THERMOMETERS SEENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY SP 250-23, LIQUID-IN-GLASS THERMOMETER CALIBRATION SERVICE.
IF NO SIGN IS GIVEN ON THE CORRECTION, THE TRUE TEMPERATURE IS HIGHERTHAN THE INDICATED TEMPERATURE; IF THE SIGN GIVEN IS NEGATIVE, THE TRUETEMPERATURE IS LOWER THAN THE INDICATED TEMPERATURE. TO USE THE CORREC-TIONS PROPERLY, REFERENCE SHOULD BE MADE TO THE NOTES GIVEN BELOW.
THE TABULATED CORRECTIONS APPLY FOR THE CONDITION OF TOTAL IMMERSION OFTHE BULB AND LIQUID COLUMN. IF THE THERMOMETER IS USED AT PARTIAL IMMER-SION, APPLY AN EMERGENT STEM CORRECTION AS EXPLAINED IN THE ACCOMPANYINGSTEM CORRECTION SHEET.
TEST NUMBER 999999P.O. NO. 99999COMPLETED 2-2-90
6.4 The National Conference of Standard Laboratories (NCSL) Ad Hoc Committee(91.3) on the Change in the International Temperature Scale that was formed to
publicize the ITS -90 and its implementation and to minimize the confusion andinconvenience of the change from the IPTS-68(75) was as follows.
Dr. B. W. Mangum, ChairmanPHY B128National Institute of Standards
and TechnologyGaithersburg, MD 20899
Mr. George W. BurnsPHY B128National Institute of Standards
and TechnologyGaithersburg, MD 20899
Dr. Stanley AbramowitzCHEM B348National Institute of Standards
and TechnologyGaithersburg, MD 20899
Dr. Richard L. AndersonInstrumentation and Controls DivisionOak Ridge National LaboratoryOak Ridge, TN 37830
Laurel Auxier, ManagerCorporate Instrumentation andMeasurement Standards
Smith Kline -Beckman,Beckman Instruments DivisionM/S H-02-C2500 Harbor BoulevardFullerton, CA 92634
Mr. Laurie H. BakerMetrology EngineeringAutonetics Marine Systems DivisionRockwell InternationalP. 0. Box 4921Anaheim, CA 92803
Mr. Jesse BerlangaNADEP/NPSD, Code 062North Island, Building 378
San Diego, CA 92135
Dr. George N. BowersDirector, Clinical LaboratoriesHartford HospitalHartford, CT 06115
Mr. Neil L. BrownWoods Hole Oceanographic InstitutionWoods Hole, Ma 02543
Dr. George T. FurukawaPHY B128National Institute of Standards
and TechnologyGaithersburg, MD 20899
Mr. Paul D. LevinePrimary Standards Laboratory0/48-75 B195ALockheed Missiles & Space CompanyP. 0. Box 3504Sunnyvale, CA 94808
Dr. Michael R. MoldoverPHY A105National Institute of Standards
and TechnologyGaithersburg, MD 20899
Mr. Henry Sostmann2307 Whitley DriveDurham, NC 27707
Mr. Sidney WeenEver Ready Thermometer Co
., Inc
.
401 Park Avenue SouthNew York, NY 10022
Mr. John WehrmeyerEastman Kodak Co.
3rd Floor, Building 69
Rochester, NY 14650-1912
Mr. James ZugelBoeing Aerospace Co.
Mail Stop 2P-30P. 0. Box 3999Seattle, WN 98124
166
7. NIST Calibration Services, Thermometers
In this appendix, calibration services offered by the Thermometry Group arespecified (taken in part from NIST SP 250).
7 . 1 Laboratory Thermometers
Technical Contact
Mailing Address:
Shipping Address
:
Jacquelyn A. WiseTel: 301/975-4822
B126 PhysicsNational Institute of Standards and TechnologyGaithersburg, MD 20899-0001
National Institute of Standards and Technology1-270 at Quince Orchard RoadA242 PhysicsGaithersburg, MD 20899-0001
Attn: J. A. Wise
Table 1. Calibration Services
Test No Items
31010C31020C31030C31040C31050C31060C31070C31080C31100C31110S31120S31130S31140S31150S31160S31200S31250S31260S
Laboratory Thermometers (0 to 150 °C)(32 to 300 °F)
Laboratory Thermometers (151 to 300 °C)(301 to 600 °F)
Laboratory Thermometers (301 to 538 °C)(601 to 1000 °F)
Laboratory Thermometers (-1 to -110 °C)(31 to -166 °F)
Laboratory Thermometers (Liquid N2 )(-196 °C or -320 °F)
Laboratory Thermometers (Liquid 2)(-183 °C or -297 °F)
Calorimetric ThermometerBeckmann ThermometerQuantity Tests of Liquid- in-Glass ThermometersSpecial Tests of Thermometers (0 to 150 °C)(32 to 300 °F)
Special Tests of Thermometers (151 to 300 °C)(301 to 600 °F)
Special Tests of Thermometers (301 to 538 °C)(601 to 1000 °F)
Special Tests of Thermometers (-1 to 110 °C)(31 to -166 °F)
Special Tests of Thermometers (Liquid N2 )(-196 °C or -320 °F)
Special Tests of Thermometers (Liquid 2 )(-183 °C or -297 °F)
Preliminary Exam or Ineligible ThermometerAdditional Copy of ReportSpecial Thermometry Services, by Prior Arrangement
Laboratory Thermometers (31010C-31100C)
This service provides for the calibration of a variety of thermometers covering the
range from -196 to +538 °C (-320 to +1000 °F)
.
Thermometers belonging to the large and varied group, which may be classed as
laboratory or "chemical" thermometers, are regularly accepted. These are of the
167
liquid- in- glass type with either solid- stem or enclosed scale. Ordinary householdor meteorological thermometers will not, in general, be accepted unless the scaleis graduated on the glass stem itself and the thermometer can be readily detachedfrom its mounting for insertion in a liquid bath. Every thermometer submitted mustbe uniquely identified by a serial number and must pass a preliminary examinationfor fineness and uniformity of graduation; for cleanliness of the mercury andcapillary bore; for freedom from moisture, gas bubbles, and cracks in the glass; foradequacy or omission of gas filling where needed; for insufficient annealing; andfor misnumbered graduations. When these or other serious defects are found, thethermometer is returned untested.
The thermometers to be calibrated are placed in a constant temperature bath alongwith a NIST-calibrated standard platinum resistance thermometer (SPRT) . The SPRTmaintains calibrations traceable to the International Temperature Scale of 1990(ITS-90) with an uncertainty of about 1 mK. (See table 2.)
Table 2: Thermometer Calibration Uncertainties
Type of Thermometer Range Uncertainty(Total Immersion)
Mercury- in- glass to 100 °C 0.03 to 0.05 °C
(graduations: 0.1-0.2 °C)
Mercury- in- glass to 300 °C 0.2 to 0.5 °C
(graduations: 1-2 °C) 300 to 500 °C 0.5 to 1.0 °C
Organic liquid-in-glass -200 to °C 0.2 to 0.5 °C
Special Tests of Thermometers (31110S-31160S)
Special tests may be conducted on temperature -measuring devices such as industrialgrade platinum resistance thermometers, digital thermometers, and thermistors.Laboratory personnel should be contacted before submitting items.
168
7.2. Thermocouples, Thermocouple Materials, and Pyrometer Indicators
Technical Contacts: George W. Burns32010C-32101C, 32150STel: 301/975-4817
Margaret G. Scroger32010C-32101C, 32150STel: 301/975-4818
Jacquelyn A. Wise32110C-32147CTel: 301/975-4822
Mailing Address: B126 PhysicsNational Institute of Standards and TechnologyGaithersburg, MD 20899-0001
Shipping Address: National Institute of Standards and Technology1-270 at Quince Orchard RoadGaithersburg, MD 20899-0001
B229]
For 32010C-32101C, 32150S[Attn: G.W. Burns or M.G. Scroger, Bldg. 221, Room
For 32110C-32147C[Attn: J. A. Wise, Bldg. 221, Room A242]
169
Comparison Calibrations , Temperature Measured with Reference Thermocouple
Test No. ItemsTemp.
TC RangeType °C
32010C S 0-1450 1
Interv. Table
32020C R 0-1450
Points
'C or 1 °F
32030C B 0-1750
32031C B 800-1750
32040C32041C32042C32043C32044C32050C
32060C
32061C
32070C
E
J
KNT
0-10000-7600-11000-11000-400
4 to 15
4 to 15
4 to 15
4 to 15
4 to 15
Min. Est.
Length Temp. Uncert(mm) (°C) (°C)
700 0-600 0.5600 to 1100 0.7
1450 2
700 to 600 0.5600 to 1100 0.7
1450 2
1000 to 600 (3 MV)600 to 1100 0.7
1450 2
1750 3
1000 800 to 1100 0.71450 2
1750 3
700 1
700 1
700 1
700 1
700 1
Comparison calibration, two point minimum, per point, for allitems aboveEach additional table of results at 1 °C or 1 °F intervals, for Type S,
R, or B at later dateEach additional table of results at 1 °C or 1 °F intervals, for Type S,
R, or B at time of testThermocouple materials tested against Pt thermoelectric standard, 4 to 15
points , 700 mm minimum lengths
Calibration at Metal Freezing Points, Minimum Diameter 0.4 mm,
Freezing Point Determination at Au, Ag, Sb and Zn
32090C 0-1450 Table, 1 °C or 1 °F 1000interv. and equationsto generate table
700 to 1100
at freezing 0.2points
to 7000.4
1450
0.3
2
32091C Type S, freezing point determination, per point, two point minimum
Calibration of Pyrometer Indicators
32100C Portable Potentiometer, first dial or range32101C Portable Potentiometer, each additional dial or range
170
Comparison Calibration of Thermocouples or Thermocouple Materials Testedagainst Pt Thermoelectric Standard, Temperature Measured with PlatinumResistance Thermometer, Minimum Length 36 Inches, Two Point Minimum
166 to 600 °F and32110C Range -110 to +300 °C and Liquid N2 (-196 °C) orLiquid N2 (-320 °F)
32120C 301 to 538 °C or 601 to 1000 °F32130C Liquid 2 (-183 °C) or (-297 °F)
Table at one degree intervals for Type T thermocouple for any of the followingoptions: (The cost of the table will be in addition to the calibration per pointcovered under fee schedule items numbered 32110C-32130C.
)
32141C Option 1: Tablepoints at -183,
32142C Option 2: Tablepoints at -183,
32143C Option 3: Tablepoints at -110, -
32144C Option 4: Tablepoints at -110, -
32145C Option 5: Table+100, +200, +300
32146C Option 6: Tablepoints at -110, -
32147C Option 7: Tablepoints at -183, -
32150S Special Tests of
from -190 to +300 °C (-310 to +572 °F) , calibration110, -50, +100, +200, +300 °C
from -190 to +100 °C (-310 to +212 °F) , calibration110, -50, +50, +100 °C
from -110 to +300 °C (-166 to +572 °F) , calibration50, +100, +200, +300 °C
from -110 to +100 °C (-166 to +212 °F) , calibration50, +50, +100 °C
to 300 °C (32 to 572 °F) , calibration points at°C
from -110 to °C (-166 to +32 °F) , calibration50 °C.
from -190 to °C (-310 to +32 °F) , calibration110, -50 °C
Thermocouples and Thermocouple Materials
Thermocouples, Thermocouple Materials , and Pyrometer Indicators (32010C-32147C)Calibration services for all commonly used types of thermocouples are providedby NIST from -196 to 1750 °C depending upon the wire or thermocouple type. Thethermocouples are calibrated by one or a combination of three general methods,depending on the thermocouple type , the temperature range , and the accuracyrequired. All three methods provide traceability to the ITS -90. In the firstmethod, thermocouples are calibrated by comparison with a standard thermocouplemaintained at NIST. In the second method, thermocouples are calibrated bycomparison with a standard platinum resistance thermometer. In the third method,thermocouples are calibrated at three defining temperatures on the ITS-90--thefreezing points of Zn, Ag, and Au, as well as at the freezing point of Sb (at
about 630.62 °C) . Below °C the thermocouple calibration is carried out in a
cryostat, while above °C stirred liquid baths, metal freezing-point cells, andelectric tube-type furnaces are employed for the calibrations. Vacuum or inertgas furnaces are also available for testing thermocouples.
Only the bare wires are required to perform the thermocouple calibrations. It
is preferable not to send insulating and protecting tubes as the rate of breakageof these in shipment is high. If the thermocouple is furnished mounted (as in
a protection tube assembly) , a nominal charge will be made for dismantling the
mounting and the various parts will be returned to the sender withoutreassembling them. Thermocouple length requirements listed in the appendix are
171
exclusive of lead wire. Lead wires (for connections) need not be sent withthermocouples. All thermocouple calibration data furnished in reports will beon the basis of a reference junction temperature of °C (32 °F) . Thecalibration results will be given in degrees Celsius or degrees Fahrenheit, as
requested by the customer.
The calibration of a thermocouple will not be undertaken if it will likely notyield the specified uncertainty or if it possesses unusual characteristics thatwould prevent the calibration or test at a reasonable cost. Only unused base-metal thermocouples and thermocouple materials will be accepted for test.
Special Tests of Thermocouples and Thermocouple Materials (32150S)
.
For requirements not covered by calibrations described above, special arrange-
ments may be made by contacting one of the specified staff members.
172
7.3 Resistance Thermometry
Technical Contacts Earl R. Pfeiffer0.65 K to 84 KTel: 301/975-4821
Gregory F. Strouse83 K to 962 °C
Tel: 301/975-4803
Mailing Address B126 PhysicsNational Institute of Standards and TechnologyGaithersburg, MD 20899-0001
Shipping Address: National Institute of Standards and Technology1-270 at Quince Orchard RoadB04 Physics,Gaithersburg, MD 20899-0001
Attn: E.R. Pfeiffer or G.F. Strouse
173
Test No, Items
33010C Capsule PRT (13.8 K to 30 °C)
33020C Capsule PRT (13.8 K to 157 °C)
33030C Capsule PRT (13.8 K to 232 °C)
33040C Capsule PRT (54 K to 30 °C)
33050C Capsule PRT (54 K to 157 °C)
33060C Capsule PRT (54 K to 232 °C)
33070C Capsule PRT (84 K to 30 °C) Ar to Ga33080C Capsule PRT (84 K to 157 °C) Ar to In33090C Capsule PRT (84 K to 232 °C) Ar to Sn33100C Capsule PRT (0 °C to 30 °C) TPW to Ga33110C Capsule PRT (0 °C to 157 °C) TPW to In33120C Capsule PRT (0 °C to 232 °C) TPW to Sn33130C Capsule PRT (234 K to 30 °C) Hg to Ga33140C Germanium and Rhodium- Iron Resistance Thermometers
(0.65 K to 24.6 K)
33150C Long Stem PRT33160C Long Stem PRT33170C Long Stem PRT33180C Long Stem PRT33190C Long Stem PRT33200C Long Stem PRT33210C Long Stem PRT33220C Long Stem PRT33230C Long Stem PRT33240C Long Stem PRT33250C Long Stem PRT33260C Long Stem PRT33270C Long Stem PRT33280C Long Stem PRT33290C Long Stem PRT33300C Long Stem PRT33310C Long Stem PRT33320C Additional Copy of Table from Results of 33010C-33310C
at Time of Test33330C Additional Copy of Table from Results of 33010C-33310C
at a Later Date33340C Minimum Charge for Unsuitable Thermometer33350S Special Tests of Resistance Thermometers33360S Special Tests of Thermometric Fixed-Point Devices33370M Measurement Assurance Program for Temperature
(8 3 K to 420 °C) Ar to Zn33380M Measurement Assurance Program for Temperature
(83 K to 661 °C) Ar to Al33390M Measurement Assurance Program for Temperature
(0 °C to 962 °C) Ar to Ag
(83 K to °C) Ar to TPW(83 K to 30 °
C) Ar to Ga(83 K to 157 °C) Ar ito In
(83 K to 232 °C) Ar ito Sn(83 K to 420 °C) Ar 1to Zn(83 K to 661 °C) Ar 1to M(234 K to 30 °C) Hg 1:o Ga(234 K to 157 °c; Hg to In(234 K to 232 °c; Hg to Sn(234 K to 420 °c; Hg to Zn(234 K to 661 °C) Hg to Al(0 °C to 30 °
C) TPW 1:o Ga
(0 °C to 157 °C) TPW to In
(0 °C to 232 °C) TPW to Sn(0 °C to 420 °C) TPW to Zn(0 °C to 661 °C) TPW to Al(0 °C to 962 °C) TPW to Ag
Resistance Thermometers (33010C-33310C)
NIST provides calibration services for standard platinum resistance thermometers
174
(SPRTs) from 13.8 K to 1235 K. Both long- stem and capsule -type SPRTs arecalibrated, providing direct access to the International Temperature Scale of1990 (ITS-90). The reproducibility of the calibration measurements of SPRTs is
better than 0.3 mK and the total uncertainty of calibration is ± 1 mK attemperatures below 661 °C. The reproducibility increases to about 0.5 mK andthe uncertainty to about ± 2 mK at 962 °C. The reproducibility of triple pointof water cells used for calibration of SPRTs is ± 50 ^K or better while that forAr, Hg, Ga, In, Sn, Zn, and Al triple-point or freezing-point cells is 0.1 mK;
and that for Ag freezing-point cells is 0.5 mK. Presently, the NIST temperaturescale in the range 13.8 K to 84 K is based on stable reference SPRTs of thecapsule type that have uncertainties of about ± 0.3 mK down to about 20 K. Below20 K, the uncertainty degrades to about ± 0.5 mK as the SPRTs become lesssensitive. For calibration of rhodium- iron resistance thermometers (RIRTs) andgermanium resistance thermometers (GRTs) in the temperature range from 0.65 Kto 24.6 K, the NIST scale is based presently on reference standard rhodium- ironresistance thermometers. The uncertainty of those calibrations is about± 0.3 mK.
To qualify for testing, either long- stem or capsule platinum resistancethermometers must meet two conditions. They must reasonably be expected to meetthe requirements of the ITS-90 for a standard interpolating instrument (i.e.,
a four lead resistor of high-purity platinum hermetically sealed in a protectingtube). Second, they must be compatible with the NIST calibration equipment.It is important that, insofar as possible, resistance thermometers be protectedfrom any mechanical shock that could alter their calibration. For shipment, the
thermometer should be softly supported within a case but not be free to rattle.
This necessitates the use of packing material that does not become compacted.The thermometer case should, in turn, be softly packed inside a shippingcontainer. The outside shipping container must be sufficiently rigid and strongthat it will not appreciably deform under the treatment usually given by commoncarriers. Styrofoam is not sufficiently rigid to be used as an outsidecontainer. Thermometers will not be returned in containers that are obviouslyunsuitable, such as those closed by nailing. Suitable containers will beprovided, for a fee, when a thermometer shipping container is not satisfactoryfor re-use.
Special Tests of Resistance Thermometers and Special Tests of Thermometric Fixed-
Point Devices (33350S and 33360S)
Special tests of various resistance thermometers and thermometric fixed pointdevices may be obtained by prior arrangement with the specified technicalcontacts
.
Measurement Assurance Program for Temperature (33370M, 33380M, and 33390M)
The purpose of this MAP service is to assure the accuracy of the calibration oftemperature standards (83 K to either 420 °C or 661 °C, or °C to 962 °C
temperature ranges) made by participating laboratories when using platinumresistance thermometry.
The MAP transport standard consists of a set of three commercial glass -sheathed
SPRTs packaged in a special shock-proof shipping container (mechanical shock or
175
sudden temperature excursions may result in shifts in calibration) . These SPRTsare used to assess both the reproducibility and the accuracy of calibrationsperformed by the participating laboratory.
MAP participants should use the techniques described in NBS Monograph 126 or NBSSpecial Publication 250-22 and the same fixed points as those used in the NISTcalibrations, or use an SPRT previously calibrated by NIST. In order to achievehigh accuracy, SPRTs used as standards should be of the matte -finish type to
avoid systematic errors arising from light-pipe effects in the glass sheath.The participant must have a triple point of water cell and a calibratedresistance bridge.
After unpacking and inspecting the SPRTs for damage, the participant shouldmeasure the resistances of the SPRTs at the triple point of water (using thetriple-point cell). This preliminary measurement of the resistances at thetriple point of water is used as a "go/no go" check to ensure that thethermometers have not been damaged in shipment.
These preliminary measurements are reported by telephone to NIST; if the valuesare consistent with the data taken by NIST before shipment, the participantshould proceed with further measurements at the fixed points of the ITS -90. NISTprovides a worksheet on which the participant should record data. Theparticipant then calculates the thermometer constants from experimental data,
records them, and prepares tables of resistance ratio or resistance versustemperature. The SPRTs are recalibrated upon return to NIST and theparticipant's data are compared with NIST's calibrations. NIST provides a plotof the participating laboratory's temperature deviation from NIST values and a
written analysis of the data, including any pertinent observations. In a typicalMAP transfer, the participant makes several measurements over a period of one
to two months. A typical turnaround time from the date NIST receives the
participant's data until a test report is sent to the participant is four to sixweeks
.
The best NIST SPRT calibrations have precisions of about 0.1 to 0.2 mK. Sourcesof error that may contribute to the total uncertainty include changes in the
calibration of the measurement instruments, changes in the SPRT itself, anduncertainty of the degree of purity of the materials used as fixed-pointreferences (e.g. , zinc). As a result of quantifying these errors, NIST currentlyassigns an uncertainty of 1 mK to the values assigned to the MAP transportstandards used below 661 °C. A standards laboratory conscientiouslyparticipating in this MAP and having suitable equipment should be able to closelyapproximate this uncertainty figure. Former participants in the MAP have haduncertainties that ranged from about 1 mK to several tens of millikelvins
.
No rigid recommendations can be given concerning how often a participant shouldutilize the temperature MAP service. Experience has indicated that whentemperature measurements are in a state of statistical control, as evidence byin-house check standards and control charts to monitor the measurement process,the participant (s) should be able to go at least three years between transfersfrom NIST without significantly degrading the confidence in the correctness of
the measurements
.
176
filST-114A
(REV.-3-89)
U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY
BIBLIOGRAPHIC DATA SHEET
1. PUBLICATION OR REPORT NUMBER
NIST/TN-12652. PERFORMING ORGANIZATION REPORT NUMBER
3. PUBUCATION DATE
August 19904. TITLE AND SUBTITLE
Guidelines for Realizing the International Temperature Scale of 1990 (ITS-90)
5. AUTHOR(S)B . W . Mangum
G . T . Furukawa6. PERFORMING ORGANIZATION (IF JOINT OR OTHER THAN NIST, SEE INSTRUCTIONS)
U.S. DEPARTMENT OF COMMERCENATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGYGAITHERSBURG, MD 20899
7. CONTRACT/GRANT NUMBER
8. TYPE OF REPORT AND PERIOD COVERED
Final9. SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (STREET, CITY, STATE, ZIP)
Same as item #6
10. SUPPLEMENTARY NOTES
DOCUMENT DESCRIBES A COMPUTER PROGRAM; SF-185, FIPS SOFTWARE SUMMARY, IS ATTACHED.
11. ABSTRACT (A 200-WORD OR LESS FACTUAL SUMMARY OF MOST SIGNIFICANT INFORMATION. IF DOCUMENT INCLUDES A SIGNIFICANT BIBLIOGRAPHY ORLITERATURE SURVEY, MENTION IT HERE.)
This Technical Note describes the International Temperature Scale of 1990 (ITS-90) that becamethe official international temperature scale on 1 January 1990, superseding the previousscales, and provides information on how the scale may be realized at different levels of
accuracy. The ITS-90 extends upward from 0.65 K, is in close agreement with the KelvinThermodynamic Temperature Scale, has much improved continuity, precision and reproducibilitythroughout its ranges over that of previous scales, and has subranges and alternative
definitions in certain ranges that greatly facilitate its use. In addition to a descriptionof the ITS-90 and how it can be realized, there are included in this document reproductions
of some articles or excerpts from articles concerned with the ITS-90. The composition of the
Comite Consultatif de Thermometrie (CCT) of the Comite International des Poids et Mesures(CIPM) at the time of the adoption of the ITS-90 is given. The differences between the
temperatures on the ITS-90 and those on the International Practical Temperature Scale of 1968,
Amended Edition of 1975, [IPTS-68(75) ] and those on the 1976 Provisional 0.5 K to 30 K
Temperature Scale (EPT-76) are tabulated. Measurement procedures for realizing the ITS-90
throughout the various ranges of the scale are given. Also, for the most important temperature
region, the region of the platinum resistance thermometer (PRT) , computational examples are
given for determining the coefficients of the relevant deviation equations for PRTs calibratedat various sets of fixed points. The effects of the introduction of the ITS-90 on electricalreference standards are addressed also.
12. KEY WORDS (6 TO 12 ENTRIES; ALPHABETICAL ORDER; CAPITAUZE ONLY PROPER NAMES; AND SEPARATE KEY WORDS BY SEMICOLONS)
fixed points; ITS-90; International Temperature Scale of 1990: SPRTs; temperature;
temperature scale; thermometers; thermometry
13. AVAILABILITY
UNLIMITED
FOR OFFICIAL DISTRIBUTION. DO NOT RELEASE TO NATIONAL TECHNICAL INFORMATION SERVICE (NTIS).
ORDER FROM SUPERINTENDENT OF DOCUMENTS, U.S. GOVERNMENT PRINTING OFFICE,WASHINGTON, DC 20402.
ORDER FROM NATIONAL TECHNICAL INFORMATION SERVICE (NTIS), SPRINGFIELD, VA 22161.
14. NUMBER OF PRINTED PAGES
190
15. PRICE
ELECTRONIC FORM i? U.S. GOVERNMENT PRINTING OFFICE: 1990 -2 6 1-913 '20673
NIST*Technical Publications
Periodical
Journal of Research of the National Institute of Standards and Technology—Reports NIST researchand development in those disciplines of the physical and engineering sciences in which the Institute
is active. These include physics, chemistry, engineering, mathematics, and computer sciences.
Papers cover a broad range of subjects, with major emphasis on measurement methodology andthe basic technology underlying standardization. Also included from time to time are survey articles
on topics closely related to the Institute's technical and scientific programs. Issued six times a year.
NonperiodicaIs
Monographs—Major contributions to the technical literature on various subjects related to the
Institute's scientific and technical activities.
Handbooks—Recommended codes of engineering and industrial practice (including safety codes) de-
veloped in cooperation with interested industries, professional organizations, and regulatory bodies.
Special Publications—Include proceedings of conferences sponsored by NIST, NIST annual reports,
and other special publications appropriate to this grouping such as wall charts, pocket cards, andbibliographies.
Applied Mathematics Series—Mathematical tables, manuals, and studies of special interest to physi-
cists, engineers, chemists, biologists, mathematicians, computer programmers, and others engaged in
scientific and technical work.
National Standard Reference Data Series—Provides quantitative data on the physical and chemicalproperties of materials, compiled from the world's literature and critically evaluated. Developed un-
der a worldwide program coordinated by NIST under the authority of the National Standard DataAct (Public Law 90-396). NOTE: The Journal of Physical and Chemical Reference Data (JPCRD)is published quarterly for NIST by the American Chemical Society (ACS) and the American Insti-
tute of Physics (AIP). Subscriptions, reprints, and supplements are available from ACS, 1155 Six-
teenth St., NW, Washington, DC 20056.
Building Science Series—Disseminates technical information developed at the Institute on building
materials, components, systems, and whole structures. The series presents research results, test
methods, and performance criteria related to the structural and environmental functions and the
durability and safety characteristics of building elements and systems.
Technical Notes—Studies or reports which are complete in themselves but restrictive in their treat-
ment of a subject. Analogous to monographs but not so comprehensive in scope or definitive in
treatment of the subject area. Often serve as a vehicle for final reports of work performed at NISTunder the sponsorship of other government agencies.
Voluntary Product Standards—Developed under procedures published by the Department of Com-merce in Part 10, Title 15, of the Code of Federal Regulations. The standards establish nationally
recognized requirements for products, and provide all concerned interests with a basis for commonunderstanding of the characteristics of the products. NIST administers this program as a supplementto the activities of the private sector standardizing organizations.
Consumer Information Series—Practical information, based on NIST research and experience, cov-ering areas of interest to the consumer. Easily understandable language and illustrations provide use-
ful background knowledge for shopping in today's technological marketplace.Order the above NIST publications from: Superintendent of Documents, Government Printing Office,
Washington, DC 20402.
Order the following NIST publications—FIPS and NISTIRs—from the National Technical InformationService, Springfield, VA 22161.
Federal Information Processing Standards Publications (FIPS PUB)—Publications in this series col-
lectively constitute the Federal Information Processing Standards Register. The Register serves as
the official source of information in the Federal Government regarding standards issued by NISTpursuant to the Federal Property and Administrative Services Act of 1949 as amended, Public Law89-306 (79 Stat. 1127), and as implemented by Executive Order 11717 (38 FR 12315, dated May 11,
1973) and Part 6 of Title 15 CFR (Code of Federal Regulations).
NIST Interagency Reports (NISTIR)—A special series of interim or final reports on work performedby NIST for outside sponsors (both government and non-government). In general, initial distribu-
tion is handled by the sponsor; public distribution is by the National Technical Information Service,
Springfield, VA 22161, in paper copy or microfiche form.
U.S. Department of CommerceNational Institute of Standards and Technology
(formerly National Bureau of Standards)
Gaithersburg, MD 20899
Official Business
Penalty for Private Use $300